The Cumulative and Emergent Automatic Deficit (CEAD) model [34, 35] explains attentional and executive difficulties in ADHD as resulting from inefficient automatization of basic processes. Such inefficiency increases cognitive load and weakens higher-order control, leading to greater interference susceptibility and reduced interference awareness. Complementary models of cognitive control deficits emphasize impairments in executive functions such as working memory, response inhibition, and set-shifting [36, 37], which limit the ability to maintain task-relevant information and suppress irrelevant input [32]. Together, these frameworks describe attentional control as a continuum: children with ADHD show pronounced deficits, those with subthreshold ADHD exhibit intermediate difficulties [29, 38], and TD peers display efficient automatization and robust executive control [39].

Based on the theoretical framework and previous empirical findings on attentional control and interference processing in ADHD, the present study aimed to examine whether children with subthreshold ADHD exhibit deficits in attentional control and interference awareness under both visual and auditory interference, comparable to those observed in children with ADHD, using an eye-tracking paradigm.

Two main hypotheses were formulated.

Hypothesis 1. Consistent with prior research documenting attentional control deficits in children with ADHD, it was hypothesized that children with ADHD would exhibit greater susceptibility to interference compared with TD peers. Specifically, under conditions involving both visual and auditory distractors, children with ADHD were expected to demonstrate poorer behavioral performance, as indexed by a lower number of correctly counted ball bounces, less efficient allocation of attention—reflected by shorter mean fixation durations on the target—and slower reorientation of gaze following distractor onset, as measured by increased latency to return to the ball. Children with subthreshold ADHD were anticipated to show a similar but attenuated pattern, exhibiting intermediate levels of performance, fixation duration, and gaze reorientation latency, suggesting that attentional control difficulties exist along a continuum from TD children to clinical ADHD.

These predictions integrate both overt behavioral measures and underlying attentional allocation processes, providing a comprehensive assessment of interference susceptibility.

Hypothesis 2. It was further hypothesized that conscious awareness of interference—the ability to recognize the influence of distracting stimuli on one’s performance—would differ significantly across groups. Children with ADHD were expected to report the distractor correctly less frequently during the 15-s pauses following each trial, reflecting limited metacognitive monitoring of attentional lapses.

Children with subthreshold ADHD were anticipated to show intermediate awareness, identifying distractors more often than clinical ADHD participants but less consistently than TD children. TD children were predicted to display the highest levels of interference awareness, indicating more effective self-regulation and monitoring of attentional performance.

Together, these hypotheses link behavioral accuracy, eye-tracking indices, and self-reported awareness, providing a multidimensional understanding of interference processing across the ADHD continuum.

One hundred and two children (mean age = 7.23 years, SD = 1.23) participated in the study: 34 with ADHD, 34 with subthreshold ADHD, and 34 TD peers, matched for age, gender, and IQ. The study employed a cross-sectional design, comparing attentional and metacognitive performance across the three groups.

Disattention and hyperactivity rating scale (SDAI). The Italian adaptation of SDAI [40] is used to screen for ADHD symptoms in school-aged children. It consists of two subscales, inattention (I) and hyperactivity (H), each comprising nine items.

Items are rated on a four-point scale ranging from 0 (never or rarely) to 3 (very often).

Total scores for each subscale range from 0 to 27, with a cutoff of 14.

Children exceeding the cutoff on the inattention subscale are classified as ADHD-I, those exceeding only the hyperactivity subscale as ADHD-H, and those exceeding both as ADHD-C. In the present sample, only ADHD-C was considered.

Psychometric evidence indicates high test-retest reliability over one month (r = 0.95) and high internal consistency (α = 0.98), supporting the scale’s reliability [40].

Aggressive behavior rating scale (SCOD). The Italian adaptation of SCOD [41] is used to assess disruptive behaviors and learning difficulties in school-aged children.

It comprises 13 items divided into a Disruptive Behavior Disorder index (8 items) and a Learning Problems index (5 items), covering both linguistic and mathematical areas.

Items are rated on a four-point scale ranging from 0 (never or rarely) to 3 (very often).

Total scores range from 0 to 24 for the Disruptive Behavior Disorder subscale (cutoff = 12) and 0 to 15 for the Learning Problems subscale (cutoff = 8).

Psychometric evidence indicates good test-retest reliability over one month (r = 0.92 for the Disruptive Behavior Disorder subscale and r = 0.89 for the Learning Problems subscale) and high internal consistency (α = 0.88 and 0.86), supporting the scale’s reliability [41].

RPM. The RPM test is a widely used tool for assessing cognitive abilities, primarily measuring fluid intelligence. Its utility stems from its relative insensitivity to socio-cultural factors, such as environmental context and educational level. The test consists of a sequence of diagrams or matrices composed of abstract geometric figures, each missing a specific element.

The standard version is divided into five sets (A to E), with increasing levels of complexity, totaling 60 items (12 per set). Participants are required to identify the underlying pattern or rule in the arrangement of figures and select the appropriate figure from the provided options to complete the series [42, 43].

Diagnostic Interview Schedule for Children (DISC-IV). The DISC-IV is a structured diagnostic interview designed to assess psychiatric disorders in children and adolescents according to DSM-5 criteria [52]. Parallel versions exist: a parent version (children aged 6–17) and a youth version (9–17 years).

In this study, the parent version of the ADHD module was used, which includes two subsections: inattention (DISC-IA) and hyperactivity–impulsivity (DISC-HI).

Each subsection includes information on the age of onset and daily functioning.

The interview evaluates the presence of sufficient symptoms across two or more settings and their impact on daily activities.

Responses are processed by a scoring program to determine whether diagnostic criteria are met, reporting symptom counts, criteria met, and functional impairment.

For this study, the impairment criterion was defined as one severe or at least two moderate impairments across six domains of daily functioning [52].

Eye-tracking. Visual attention and reaction times were assessed using a Tobii Pro X3-120 remote eye tracker. This device records eye movements non-invasively, allowing natural interaction with stimuli [53]. It has a sampling rate of 120 Hz and monocular accuracy of approximately 0.34°, providing reliable measurements of fixations, saccades, and reaction times in brief visual tasks.

Individual calibration was performed using a nine-point protocol. Eye-tracking data were synchronized with experimental videos and distractors, enabling assessment of attentional control, selective attention, and distractor awareness [54].

Task. Children’s attentional performance was assessed using a video-based eye-tracking task adapted from previous studies on attentional control in children with ADHD [55, 56]. Each participant was tested individually in a quiet, distraction-free room and seated comfortably in front of the display. Prior to data collection, the eye tracker was calibrated to ensure accurate gaze tracking. The task consisted of three short video clips depicting a bouncing ball, corresponding to three experimental conditions: no interference, visual interference, and auditory interference (Figure 1). Each trial contained 20 bounces, with each bounce lasting 1.5 s, for a total trial duration of 30 s. After each trial, a 15-s pause was introduced, during which the experimenter asked children whether they had noticed any changes in the stimuli, providing an index of distractor awareness (Figure 2).

In the visual interference condition, distractors included changes in background color, ball color, variations in ball size, and appearance of objects.

In the auditory interference condition, distractors included sounds such as a dog barking, a glass breaking, a bell ringing, and a whistle.

The no interference condition presented the ball bouncing without any additional visual or auditory stimuli.

Each participant completed one trial per condition, with the order of conditions randomized across participants. Eye-tracking data were collected throughout each trial to measure fixation duration on the ball, latency to return gaze to the ball after distractor onset, and overall attentional allocation.

Latency to return gaze to the ball after distractor onset represents the efficiency of attentional reorientation following distraction, measured from the onset of the distractor to the first fixation on the target. Distractors were presented for 500 ms, a duration consistent with previous studies on attentional reorientation in children [57], which reported mean gaze return latencies in the range of 400–750 ms under similar interference conditions.

This design allowed for the simultaneous assessment of behavioral performance (number of correctly counted bounces) and both implicit and explicit awareness of distractors.

The study was conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from the parents or guardians of all participants prior to data collection, which took place between February and May 2024.

All assessments were administered by a trained experimental psychologist during school hours (9:00 a.m. to 12:00 p.m.). Before beginning the experimental task, the examiner verified that all children could count to 100 and accurately track moving objects. Participants then received standardized instructions, and the eye-tracking system was individually calibrated. Following calibration, participants completed the three video trials according to the protocol described above. The order of the three conditions was counterbalanced across participants using a Latin-square design to control for potential order and fatigue effects.

After each trial, a brief post-trial interview assessed whether participants noticed any changes or unusual events, providing additional information on conscious detection of distractors. All procedures were standardized to ensure consistency across participants and to minimize potential confounding variables. The total session, including instructions, calibration, task completion, and pauses, lasted approximately 5 min per participant.

Behavioral and eye-tracking data were analyzed using IBM SPSS Statistics 25.

To test Hypothesis 1 (interference susceptibility) and explore attentional allocation, a 3 (group: TD, subthreshold ADHD, ADHD) × 3 (condition: no, visual, auditory interference) mixed-design ANOVA was conducted separately for each dependent variable: the number of correctly counted ball bounces, mean fixation duration on the ball (ms), and latency to return gaze to the ball after distractor onset (ms). Group was treated as a between-subjects factor, and condition as a within-subjects factor. Sphericity was assessed using Mauchly’s test, and Greenhouse-Geisser corrections were applied when necessary. Significant main effects and interactions were followed up with Bonferroni-corrected pairwise comparisons. Effect sizes were reported using η²_p for ANOVA and Cohen’s d for post hoc comparisons. To address potential type I error inflation due to multiple correlated dependent variables (e.g., behavioral and eye-tracking measures), a multivariate ANOVA (MANOVA) was conducted as a sensitivity analysis.

The dependent variables included the number of correctly counted bounces, mean fixation duration, proportion of trial time fixating the target, and latency to return gaze to the target following distractor onset. Multivariate effects were assessed using Wilks’ λ, and significant effects were followed by univariate ANOVAs with Bonferroni correction to examine the contribution of individual dependent variables. To examine Hypothesis 2 (interference awareness), the number of children correctly detecting distractors during the 15-s pause after each trial was analyzed using chi-square tests of independence. Pairwise comparisons with Bonferroni correction were conducted to identify differences between groups.

All statistical tests were two-tailed, and statistical significance was determined using a p-value threshold of p < 0.05; Bonferroni-adjusted p-values were used for post hoc comparisons where applicable. Descriptive statistics (mean and SD) were reported for all dependent variables. When appropriate, post hoc tests were accompanied by effect sizes (Cohen’s d) to quantify the magnitude of group differences [53]. Eye-tracking data were preprocessed to remove blinks and invalid samples. Mean fixation duration on the ball was computed for each trial, and the proportion of the trial spent fixating on the ball was calculated as the total fixation time on the ball divided by the total trial duration, expressed as a percentage.

Latency to return gaze to the ball after distractor onset was measured from the onset of the distractor to the first fixation on the ball, exceeding 100 ms. These measures provided convergent evidence regarding attentional allocation and reorientation efficiency across groups and conditions.

Behavioral results demonstrated that ADHD children counted fewer ball bounces under interference conditions compared with both TD and subthreshold ADHD peers, whereas subthreshold ADHD children performed better than ADHD children but worse than TD children. Eye-tracking metrics revealed that ADHD children had shorter fixation durations on the target and longer latencies to return gaze following distractor onset, indicating reduced attentional stability and slower reorientation. Subthreshold ADHD children exhibited intermediate performance. These results provide robust evidence that attentional control deficits are present not only in clinical ADHD but also in subthreshold populations, supporting a dimensional rather than categorical view of cognitive control deficits [16, 29, 38].

Assessment of interference awareness further showed graded differences across groups. TD children most frequently detected distractors, ADHD children the least, and subthreshold ADHD children fell in between. This pattern suggests that deficits in metacognitive monitoring co-occur with attentional control impairments, consistent with prior studies showing reduced awareness of distractors in ADHD [31, 33]. Importantly, the partial awareness observed in subthreshold ADHD children indicates that even subclinical attentional difficulties can influence performance and everyday functioning, emphasizing the functional relevance of subthreshold symptomatology.

These findings align with and extend previous literature on ADHD and attentional control. Prior research has consistently reported heightened susceptibility to interference and slower attentional recovery in ADHD [25–28]. The intermediate performance of subthreshold ADHD children supports earlier suggestions that subclinical ADHD symptoms are associated with measurable cognitive difficulties, including deficits in executive control and attentional regulation [8, 9, 12]. Eye-tracking evidence in the present study provides direct, real-time confirmation of these impairments, demonstrating that both overt behavioral outcomes and underlying attentional allocation mechanisms are affected. The results are also consistent with theoretical frameworks integrating attentional encoding and executive control deficits. The CEAD model [34, 35] posits that early inefficiencies in attentional encoding disrupt the automatization of basic cognitive processes, overloading higher-order control and increasing distractibility. Our findings suggest these processes operate along a continuum, with subthreshold ADHD representing a moderate level of inefficiency that produces observable but less severe deficits compared with clinical ADHD. Cognitive control models emphasizing working memory, response inhibition, and set-shifting [36, 37] are also supported, as interference susceptibility and delayed gaze reorientation reflect limitations in maintaining task-relevant representations and suppressing irrelevant stimuli. Beyond confirming previous evidence, the present results refine both the CEAD and cognitive control models. Within the CEAD framework, our data suggest that deficits in automatization and attentional encoding are graded: subthreshold ADHD children exhibit partial automatization of attentional routines, leading to moderately increased cognitive load and interference susceptibility. This finding highlights a potential compensatory mechanism, whereby partial awareness of distraction supports adaptive regulation despite inefficient automatization. From the perspective of cognitive control theories, the current results indicate that executive inefficiencies (e.g., in inhibition and working memory) interact dynamically with perceptual encoding processes. This interplay demonstrates that attentional stability and metacognitive awareness are functionally linked rather than independent domains of deficit.

These findings have important implications for early identification and intervention. The demonstration that subthreshold ADHD is associated with measurable attentional and metacognitive deficits suggests that early support should not be limited to children meeting full diagnostic criteria. Interventions could include computerized attention training programs aimed at strengthening sustained and selective attention, mindfulness-based practices to enhance awareness of distraction and emotional regulation, and metacognitive therapy approaches focusing on self-monitoring, performance feedback, and strategic attentional shifting.

These methods may benefit children across the ADHD continuum by improving both automatic and controlled aspects of attention, potentially reducing the accumulation of functional difficulties over time.

Several limitations should be acknowledged. The sample included only children aged 6–8 years, limiting generalizability to older age groups or adolescents. Interference awareness was measured using post-trial self-report, which may not fully capture real-time attentional monitoring. Future research should employ continuous measures, such as neurophysiological indices (e.g., electroencephalography, pupillometry) or concurrent eye-tracking metrics, to better assess attentional dynamics. Longitudinal studies are needed to determine whether subthreshold ADHD reflects a transient developmental stage, a stable intermediate phenotype, or a risk factor for progression to full ADHD. Replication across multiple cognitive tasks and in diverse cultural contexts would strengthen the generalizability of the findings. Another limitation concerns the cultural and contextual characteristics of the sample. All participants were recruited from primary schools in southern Italy, a context that may differ in educational practices, attentional demands, and classroom environments compared with other cultural settings. As cultural factors can influence attentional regulation and metacognitive development, cross-cultural replications are essential to determine whether the observed continuum of attentional control and interference awareness generalizes to children from different socio-educational systems.

In summary, the present study demonstrates that both ADHD and subthreshold ADHD are associated with deficits in attentional control and reduced interference awareness, with subthreshold ADHD children showing intermediate performance. These results support a continuum model of cognitive control impairments, highlighting the functional relevance of subclinical symptoms and the need for early detection and interventions targeting both attentional and metacognitive processes across the ADHD spectrum. While our findings underscore the presence of a graded continuum of attentional control and awareness, this perspective should be viewed as complementary to established diagnostic frameworks rather than as a substitute. Categorical classifications continue to serve critical functions in clinical assessment, diagnosis, and treatment planning.