Descriptive statistics are available in Table 1. A total of 158 participants took part in the study, of whom 6 were excluded due to incomplete task or questionnaire data. The final sample consisted of 152 participants, divided into four groups (n = 38 per group): individuals with a clinical diagnosis of ADHD or clinical diagnosis of ASD, and their matched non-clinical counterparts. Of these, 88 identified as female, 59 as male, and five as non-binary. The majority of participants (60%) identified as White European/British/Irish with a college-level education, while the remaining 40% represented diverse ethnic backgrounds, including Asian, Black, and Mixed Ethnicity. Ages ranged from 18 to 57 years. The non-clinical cohort were recruited from the University of Sheffield and the surrounding community. For consistency with prior literature, we use the term ‘non-clinical’ to refer to participants without a formal clinical diagnosis of ADHD or ASD. This terminology is used operationally to indicate diagnostic status only and does not imply diagnostic negativity.

Participants in the clinical cohort were recruited through National Health Service (NHS)-affiliated support services groups and NHS-connected local councils’ neurodiverse groups for individuals with ADHD and ASD. Participants had an existing clinical diagnosis of ADHD or ASD established prior to participation and evidence of clinical diagnosis was provided at enrolment. Consistent with standard practice in adult ADHD and ASD research, participants were not re-assessed diagnostically for the purposes of the study, and clinical diagnosis was treated as the baseline criterion for inclusion in the clinical cohorts (41). Based on scores on standardized measures used in this study (see below), participants from the ADHD clinical cohort were predominantly of the inattentive phenotype, followed by combined and hyperactive phenotypes, in line with other research within adult populations (42, 43), while the ASD clinical cohort scored highest on difficulties within the social communication domain, followed by repetitive behaviors, in line with Leford-Besnard et al. (44). In the clinical ADHD cohort, 20 participants reported current use of ADHD-related medication. In contrast, none of the ASD clinical cohort or any other participants reported psychotropic medication use. Finally, comorbid conditions such as anxiety, depression and learning disabilities were reported by many participants across the clinical cohorts and are detailed in Supplementary Table S1. These were not modelled as independent variables, as contemporary transdiagnostic frameworks (e.g., HiTOP, p-factor models [45, 46]) conceptualize comorbidity as arising from shared latent dimensions rather than from independent, separable disorders. Within this perspective, comorbidities are integral to the clinical phenotype and diagnostic process in adult ADHD and ASD, rather than nuisance variance to be controlled.

All participants provided informed consent and received detailed participant information and debrief forms, in accordance with the Declaration of Helsinki (47). Ethical approval was obtained from the University of Sheffield Ethics Committee. Each participant was assigned a unique, anonymized code to maintain confidentiality. This study was preregistered on the Open Science Framework (OSF) prior to data analysis and a link to it can be found in the Preregistration statement below.

To match the groups in terms of their scores using standardized tools, the presence of ADHD and ASD traits were assessed using two self-report measures specifically designed to assess dimensional psychiatric traits in adults, administered via the online platform Qualtrics XM (Qualtrics, U.S.A.) and presented in a randomized order. ADHD traits were assessed using the Adult ADHD Self-Report Scale (ASRS; 37), and ASD traits were measured using the Broad Autism Phenotype Questionnaire (BAPQ; 48).

The ASRS is an 18-item instrument based on the DSM-IV diagnostic criteria for ADHD, comprising two subscales—Inattention (IN) and Hyperactivity/Impulsivity (HP)—each containing nine items. Responses are rated on a five-point Likert scale (Never, Rarely, Sometimes, Often, Very Often) indicating how frequently each statement applied to participants over the past six months (e.g., IN: “How often do you have problems remembering appointments or obligations?”; HP: “How often do you interrupt others when they are busy?”). In the current study Cronbach’s alpha was above.87.

The BAPQ consists of 36 items divided into three subscales—Aloof Personality (AP), Pragmatic Language (PL), and Rigid Behavior (RB). Participants rated how characteristic each statement was of them during the preceding six months using a six-point Likert scale (Very Rarely, Rarely, Occasionally, Somewhat Often, Often, Very Often). Example items include: AP – “I would rather talk to people to get information than to socialize,” PL – “It’s hard for me to avoid getting sidetracked in conversations,” and RB – “I am comfortable with unexpected changes in plans.” In the current study Cronbach’s alpha was above.70.

Visual orientation discrimination thresholds were measured using a method of constant stimuli combined with a two-alternative forced-choice (2AFC), adaptive, and randomly interleaved staircase procedure, following the one-up–three-down design described by Dickinson et al. (29) and Edden et al. (49). On each trial, participants viewed a reference sinusoidal grating followed by a target grating, each presented for 350 ms and separated by a 500 ms interstimulus interval. Their task was to indicate whether the target grating was rotated clockwise or anticlockwise relative to the reference grating by pressing the right or left arrow key, respectively (see Figure 1 for trial sequence).

Stimuli were generated using PsychoPy v2022.2.4 and presented in a mixed sequence, such that participants were never exposed exclusively to vertical or oblique orientations. Each grating subtended 4° of visual angle (4 cm diameter), had a spatial frequency of three cycles, mean luminance of 45 cd/m², and 80% contrast. Stimuli were displayed on a linearized LCD monitor viewed through a circular aperture to eliminate orientation cues from the screen edges. A chinrest positioned 57 cm from the monitor stabilized head position. Gamma and luminance calibration were performed using PsychoPy’s gamma correction utility and a photometer.

The target grating initially differed from the reference by 5° and adapted according to performance: the orientation difference decreased after three consecutive correct responses (increasing task difficulty) and increased after one incorrect response (decreasing task difficulty) until nine reversals occurred. After each reversal, the step size was reduced to 75% of its previous value. Orientation differences ranged from 0.001° to 20° relative to the reference. The task comprised two conditions: a vertical condition (reference = 0°) and an oblique condition (reference = 45°). For each condition, two staircases tracked clockwise and anticlockwise rotations, resulting in four interleaved staircases (0°clockwise, 0° anticlockwise, 45°clockwise, 45° anticlockwise). Staircases converged on 79% correct performance (50).

An information sheet was provided, and written consent was obtained from all participants prior to beginning the study. Participants completed the battery of questionnaires up to 24 h before attending the testing session for the visual task. Upon arrival, they were asked to store away any light-emitting devices and remained in the testing room for 30 min to allow for dark adaptation, during which all light sources were switched off and monitor covers were placed in position [following Kalloniatis & Luu, (51)]. The visual task comprised two main sections: a practice session, followed by a 40 s rest period, and the main task, which was divided into blocks with 2 min breaks between them and lasted approximately 17 min in total. After completing the task, participants were debriefed and given the opportunity to ask questions about the study.

The differential performance of clinical and non-clinical ASD participants suggests that those individuals with a clinical diagnosis may be qualitatively different to those without, in spite of their match in terms of their standardized clinical characteristics. As we mentioned in the introduction, a clinical diagnosis is a considerably more comprehensive evaluation of a person’s symptomology, requiring a detailed synthesis of symptom presentation, clinical interview, developmental history, and functional impairment, often integrating information from standardized tools along with direct observation, informant reports, and clinician judgment (36, 37). Such a comprehensive evaluation will potentially detect disorder specific characteristics that are not encapsulated in commonly used report tests like the BAPQ. The fact that clinical and non-clinical ADHD participants differed only in terms of the severity of the deficit, in contrast to the clinical ASD participants who outperformed the non-clinical ASD participants, suggests that the two conditions may differ in how clinical diagnosis modulates visual perceptual processing, even when trait severity is held constant.

These disorder-specific trajectories are consistent with two previous studies which researched the extent to which sensory processing is comparable in the two conditions (26, 54). In the first study, the two conditions formed distinct interpersonal and intrapersonal clusters of behaviors in a network analysis: ADHD traits were linked to inattention and difficulties maintaining a train of thought, whereas ASD traits were uniquely associated with social anxiety—often underpinned by an increased focus on detail over context (55). Similarly, in the second study, non-clinical cohorts with ADHD and ASD traits were tested on a VOD task. Although both groups showed similarly impaired performance (elevated thresholds), visual performance was linked to anxiety in the ADHD cohort but not the ASD cohort, again distinguishing the two conditions. These distinct profiles are supported by evidence that ADHD-related impairments arise from alterations in sustained attention and working memory (56, 57), whereas impairments in ASD appear to reflect a cognitive style characterized by heightened sensitivity to fine-grained features and subtle differences in stimuli, but reduced ability to integrate these into a global whole (58–60).

An important question is the extent to which our findings have implications for the idea that ADHD and ASD are dimensional disorders. At a descriptive level, our results are broadly consistent with the notion that ADHD aligns more closely with a dimensional account, insofar as the clinical group showed greater impairment than the non-clinical group while differing primarily in degree rather than in kind. This pattern is compatible with models in which ADHD traits are continuously distributed, with measures reflecting a broad factor structure and clinical and non-clinical individuals occupying different positions along that continuum (see 61–63). At the same time, the nature of this dimensionality remains ambiguous. This is particularly relevant given that the ASRS does not capture several clinically salient features of ADHD that may contribute to group differences. In contrast, the pattern observed for ASD appears less readily accommodated by a simple dimensional account. Clinically diagnosed participants demonstrated superior performance relative to high-trait, non-clinical individuals, and this finding complements prior evidence of a negative association between autistic traits and visual discrimination thresholds in non-clinical samples (26). Taken together, these observations suggest that increasing autistic trait levels do not map monotonically onto sensory performance in the same way across clinical and non-clinical groups. Rather than implying a strictly categorical distinction, this pattern is consistent with models proposing a hybrid structure for ASD, in which continuously distributed traits coexist with qualitative clustering of core features—particularly within social-communication, repetitive behavior, and sensory domains—that distinguish clinically diagnosed individuals from high-trait non-clinical cases (64). From this perspective, a clinical ASD diagnosis may reflect a distinct configuration and organization of traits and functional impacts, rather than simply a higher position on a single continuum of autistic characteristics. Consequently, approaches that integrate both dimensional and categorical perspectives may be better suited to capturing the heterogeneity and boundary structure of ADHD and ASD, as well as the risk of overlooking atypical or subthreshold presentations when relying on either framework alone (65, 66).Further research across all sensory modalities is needed to ascertain nosological patterns.

The finding that clinical groups in both conditions performed differently to non-clinical groups bring us back to the question of the impact of a clinical diagnosis, which appears to depend on more than just passing a certain threshold score of severity. Our results within the context of visual sensory performance and more specifically the ability to visually discriminate in ADHD and ASD, suggest that there seems to be a qualitative component to diagnosis which captures not just severity of traits but the impact of core aspects of the disorder on daily life which are unique for the clinical participants. In recent years issues with self-reported and self-diagnosed ADHD and ASD have been raised by clinical researchers (such as 67–69) in line with the notion that it is possible unique trajectories and patterns of behavior are observed in both self-diagnosed and clinician diagnosed cohorts. Banker et al. (67, 70) report a significant lack of agreement between the self-rated and clinician-assessed symptoms in groups of clinical and self-rated ASD cohorts, suggesting symptoms reported in each of the two groups should be interpreted separately rather than as gradient of magnitude. Further research is needed to understand whether and how clinical diagnosis could differentiate ADHD and ASD across cognitive or sensory domains.

The unique formulation of impact coming from a clinical diagnosis highlights that diagnosis goes beyond what is revealed by simply increasing or decreasing trait severity scores. Clinicians, as discussed above, use multidimensional assessment focusing on functional impact and nuanced symptoms, which allows for the identification of distinctive features in clinical cases. Whilst the BAPQ may capture traits like aloofness, rigidity, and pragmatic language issues, it does not adequately detect the subtle qualitative differences in social reciprocity, nonverbal communication or adaptive/masking behaviors that can characterize ASD (71), particularly in adults and which play a crucial role in the diagnostic process (67). Clinical interviews and comprehensive assessments integrate these missing dimensions via direct observation, collateral information, real-world history, and probing for contextual factors. This richer, qualitative, multi-informant evaluation allows clinicians to identify the presence, configuration, and impact of symptoms beyond what questionnaires alone can capture, ensuring a more accurate and ecologically valid diagnosis. Such assessment in relation to ASD might allow diagnostic processes to systematically differentiate people whose local bias and enhanced sensory discrimination are core, enduring aspects of their functioning from those who merely score high on questionnaires but lack the persistent, generalizable pattern or real-world impact needed for a clinical diagnosis.

Likewise, in relation to ADHD, the ASRS primarily assesses the core symptoms - inattention, hyperactivity, and impulsivity, however, it misses key aspects crucial to the configuration of the condition, such as emotional dysregulation (e.g., irritability, mood swings, frustration intolerance), difficulties with time management and organization, and variability across contexts, all of which are often clinically significant and may contribute as much, if not more, to daily impairment as the core ADHD features but are yet under-recognized by the scale alone (72–74).

While previous findings concerning low level perception in ADHD are scarce, research on ASD and sensory processing has produced conflicting results in the context of VOD. As discussed above, some previous research using visual discrimination performance found that ASD cohorts perform better with increases in severity (28, 29) but other researchers found either no relationship between performance and severity (30) or a decrease in performance with increased severity (26). Although contradictory, the results of the current study may offer insight into the basis of the contradiction. We have found that individuals with a clinical diagnosis have enhanced VOD performance, which is also supported by the findings of Bertone et al. (28). The contradictory finding come from studies that have used participants they’ve classed as ‘non-clinical’. One possibility, again suggested by our findings, is that for a cohort that is truly non-clinical (i.e. would not be diagnosed as ASD by a clinician), VOD performance is impaired as in our earlier work (26). However, other studies using ostensibly non-clinical cohorts may have inadvertently recruited individuals who would have attracted a clinical diagnosis under the correct circumstances. The extent to which the tested sample contained such individuals could turn impaired performance into enhanced performance (as in 29) or simply offset impaired performance to arrive at overall neutrality (as in 30).

In view of the points discussed above, the major contribution of our findings is that they demonstrate - in spite of high level similarities in sensory processing abnormalities in ASD and ADHD - the conditions differ at a more granular level of analysis. Although ASD and ADHD can be conceptualized as related in their sensory impairments, they do not seem to be identical. Within the visual sensory domain examined here, ADHD and ASD showed different patterns of diagnostic modulation on this orientation discrimination measure that are consistent with dimensional and hybrid interpretations, respectively, though investigation over various aspects of sensory processing and across all modalities is needed to understand more definitively disorder specific structures. We cannot therefore postulate that they arise from a common neural substrate, at least in terms of sensory processing component to the disorders. We also demonstrate that the impact of a clinical diagnosis is far more nuanced than simply representing a point of higher severity on a continuous trait scale; rather, clinical diagnosis identifies qualitative and core differences in symptom profiles, functional impact, and neurocognitive organization that are not captured by rating scores alone.

Although we assess sensory processing within the visual modality in this study, we did not consider sensory differences in other modalities, which may diverge in patterns and relevance. This is important in order to ascertain the full picture of sensory functioning. In addition, as some research has suggested different profiles of impairment in ADHD between adolescent and adult cohorts (75), it is important to look at younger ages and ascertain whether the same clinical impact on behavior will hold. Further, this study does not discuss the impact of comorbidities. The important impact comorbidities have on sensory processing has already been described in non-clinical cohorts of ADHD and ASD (26) and an assessment of their impact in clinical populations is needed in adults, however this is beyond the scope of the current paper. Finally, conclusions regarding dimensional versus categorical models are necessarily limited by the use of a single perceptual task within one sensory modality and should be interpreted within the context of the current research. Larger cohorts and examination across all modalities is to needed to evaluate these dimensional–hybrid interpretations more robustly.