The relationship between attention and intelligence is a central issue in cognitive neuroscience, as both domains rely on overlapping neural substrates and cognitive mechanisms [1–4]. Attention is a basic condition for all higher functioning and for any kind of intellectual activity. The ability to selectively focus on relevant stimuli, while simultaneously inhibiting distractors, underlies the overall process, from primary stimulus encoding to higher-order cognitive processing, especially if we consider that the human capacity to process information is rather limited compared to the very large amount of internal and external stimuli [5–7].

Visuospatial attention allows for the selective processing of spatially organized visual information, the extraction of rules and patterns, and the concentration on specific visual stimuli while suppressing irrelevant details [8].

Within the Cattell-Horn-Carroll theory of intelligence, visuospatial attention plays a significant role in problem-solving and fluid intelligence tasks by enhancing visual processing, supporting visual working memory, and maintaining attentional focus. Similarly, the process overlap model posits that attentional control, including visuospatial attention, is a core mechanism underlying performance on intelligence tasks [9, 10].

Accordingly, stronger visuospatial attention skills predict better outcomes in problem solving and nonverbal reasoning. Eye-tracking studies show that children with higher intelligence scores allocate attention more strategically, concentrating on relevant features rather than scanning randomly [11–13].

Developmental research further suggests that these associations are particularly strong during childhood, when attentional systems are still maturing and exert a strong influence on reasoning ability [14–16].

At the neuroanatomical level, both visuospatial attention and fluid intelligence depend on frontoparietal networks subserving executive control, working memory, and abstract reasoning, particularly the posterior parietal cortex, dorsolateral prefrontal cortex, and inferior frontal gyrus [17, 18].

Prefrontal structures play a pivotal role by integrating attentional processes with higher-level executive functions. Because attention pervades virtually all aspects of cognition, deficits in attentional development can compromise encoding, storage, retrieval, reasoning, and planning [1, 19–22].

Consequently, low academic performance may stem from difficulties in sustaining attention or maintaining working memory [23–28].

Nevertheless, poor scores on intelligence tests are often considered exclusively as evidence of global cognitive impairment. Raven’s Colored Progressive Matrices (RCPMs), for instance, are widely used as a measure of nonverbal fluid intelligence and are evaluated as a single measure of general intelligence. However, several studies have shown that performance depends on distinct cognitive abilities, such as analytic reasoning, analogical reasoning, and figural processing, and recruits different neural substrates, including bilateral frontal and left temporo-parieto-occipital regions. More recent neuropsychological research has identified three qualitative item clusters: (i) items requiring complex analytical and gestalt reasoning; (ii) items requiring abstract analogical reasoning; and (iii) items requiring simple perceptual completion based primarily on visuospatial attention [29–32]. This evidence indicates that similar overall RCPMs scores may result from different underlying cognitive strategies, including attentional mechanisms. Attention deficit hyperactivity disorder (ADHD) is a neurodevelopmental condition characterized by impairments in attentional control, manifesting as inattention, hyperactivity, impulsivity, or their combination [33]. Attention deficits, particularly difficulties in sustaining and shifting spatial attention, are more prominent in the inattentive subtype of ADHD (ADHD-I), while deficits in impulse control during visuospatial tasks are more strongly associated with the hyperactive subtype (ADHD-H). Such deficits affect learning, memory, and executive functioning, with broad consequences for academic achievement and social adjustment [34–39]. The present study investigates visuospatial attention and both verbal and nonverbal intelligence in children with the hyperactive subtype of ADHD. We hypothesized that poor performance on tasks of fluid intelligence in this group would primarily reflect deficits in attentional control rather than global intellectual impairment. Specifically, compared to typically developing peers, children with ADHD-H were expected to perform worse on a visuospatial attention task [the Visual Form Discrimination Test (VFDT)] and on RCPMs, which is heavily dependent on visuospatial processing, but not on a verbal abstraction task. This study thus aims to contribute to a more differential assessment of cognitive abilities in ADHD, clarifying the extent to which attentional deficits influence measures of intelligence. Such differentiation is critical in both clinical and educational settings, as it allows practitioners to distinguish poor attentional performance from poor general intelligence and to address each domain appropriately in rehabilitation.

No significant differences emerged between groups with respect to socioeconomic status or parental education levels (all p values > 0.05) (Table 1).

Group comparisons are summarized in Table 2. Children in the control group performed significantly better than children with ADHD-H on both visuospatial and nonverbal reasoning tasks. Specifically, controls obtained significantly higher scores than children with ADHD-H on the VFDT (t₁₃₆ = 7.97, p < 0.001, d = 1.35, 95% CI [0.98, 1.72]) and on the RCPMs (t₁₃₆ = 4.56, p < 0.001, d = 1.69, 95% CI [1.30, 2.08]). These large effect sizes indicate substantial group differences in visuospatial attention and fluid intelligence. No significant group difference was observed for verbal abstraction abilities (t₁₃₆ = 1.59, p = 0.116, d = 0.26, 95% CI [−0.08, 0.60]), suggesting that verbal reasoning remains relatively preserved in children with ADHD-H.

To examine this pattern qualitatively, the VFDT was divided into two halves (Items 1–8 and Items 9–16). The error distribution in the VFDT revealed a clear trend in performance across the test sequence. A repeated-measures ANOVA with Set (first half vs. second half) as the within-subject factor and Group (ADHD vs. control) as the between-subject factor revealed a significant main effect of Set [F_(1,136) = 94.6, p < 0.000001], indicating that performance declined significantly in the second half of the test. The number of correct responses was significantly higher in the first half compared to the second (p < 0.000001). A descriptive trend suggested a greater performance decline across task halves in the ADHD-H group compared to controls, suggesting that attentional control and sustained focus deteriorate as the visuospatial task progresses over time in children with ADHD-H. Peripheral errors were also analyzed across the two sets and showed significant differences in all groups, being more frequent in the second half sets.

Furthermore, since correct responses in the test are evenly distributed across the four quadrants (4 correct responses for each quadrant: above, below, right, left), a qualitative analysis was conducted to examine whether the spatial position of the correct alternatives on the response sheet influenced performance. When comparing items with correct answers positioned on the left (options 1 and 3) vs. the right side (options 2 and 4), a significant main effect of position emerged [F_(1,136) = 46.4, p < 0.000001]. Correct responses were significantly more frequent when the correct alternative appeared on the right side of the page. Similarly, when the distribution of correct responses was analyzed according to their vertical position, top (options 1 and 2) vs. bottom (options 3 and 4), the main effect of Side was not significant [F_(1,136) = 0.04, p = 0.85]. A non-significant descriptive trend indicated a higher number of correct responses for top-positioned alternatives in the ADHD-H group.

Table 3 reports the mean RCPM scores for both groups, including the total score (overall number of correct responses out of 36) and the three cognitive components identified in previous factor-analytic studies: similarity, gestalt reasoning, and analogy. Children in the control group obtained higher and more homogeneous scores (with smaller standard deviations) than those with ADHD-H across all parameters. However, the difference between groups on the similarity factor (11 items requiring perceptual matching by visual similarity) was not statistically significant (Controls: mean =10.75 ± 0.48; ADHD-H: mean = 10.13 ± 2.87; p > 0.05).

In contrast, significant differences emerged in the remaining components. On the gestalt reasoning factor (15 items requiring complex integration of visual information), the control group outperformed the ADHD-H group (mean = 12.98 ± 2.78 vs. 10.89 ± 3.95; t₁₃₆ = 2.09, p < 0.05). A similar pattern was found for the analogy factor (10 items requiring abstract reasoning by analogy), where the difference was even larger (mean = 4.56 ± 1.54 vs. 2.10 ± 3.98; t₁₃₆ = 2.46, p < 0.05).

Finally, the difference in total RCPMs score was highly significant, with the control group obtaining markedly higher scores (mean = 27.58 ± 1.81) than the ADHD-H group (mean = 23.02 ± 3.31; t₁₃₆ = 4.56, p < 0.001).

Table 4 presents the mean scores of the two groups on the Verbal Abstraction Test. Overall, children in the control group obtained slightly higher mean scores than those with ADHD-H, both in the total score and in the two individual subtests. Specifically, control participants performed marginally better on the Comprehension subtest (mean = 4.65 ± 0.25 vs. 4.32 ± 2.31) and on the Verbal Absurdities subtest (mean = 7.40 ± 0.45 vs. 7.12 ± 2.23). Similarly, the total score tended to be higher in the control group (mean = 11.32 ± 0.11) compared to the ADHD-H group (mean = 10.91 ± 2.25).

However, none of these differences reached statistical significance (p > 0.05), indicating that verbal abstraction abilities were largely comparable between the two groups.

The present study investigated visuospatial attention, nonverbal fluid intelligence, and verbal abstraction in children with ADHD-H, compared with typically developing peers. The results provide evidence that attentional control plays a fundamental role in nonverbal reasoning performance and may underlie poorer performance on nonverbal reasoning tasks in ADHD-H, while verbal abstraction abilities remain relatively intact.

Children with ADHD-H scored significantly lower than controls on the VFDT, with large effect sizes, indicating a pronounced deficit in visuospatial attention during discrimination of complex visual forms. The analysis of both central and peripheral figures was significantly impaired, with a higher number of errors, suggesting difficulties in selectively exploring and integrating visuospatial exploration, flexible visual search strategies, and sustained attentional maintenance to match target figures with one of several alternatives.

Overall, the performance pattern is consistent with inefficient visuospatial scanning and reduced attentional control during complex visual processing.

Furthermore, in the second half of the VFDT, the ADHD-H group showed a greater decline in performance compared to the first half, a pattern that was more evident at a descriptive level than in the control group. This finding suggests increasing difficulty in maintaining visuospatial attention and attentional control as task demands persist over time. Importantly, this effect cannot be attributed to differences in item difficulty, since the VFDT does not involve a progressive increase in complexity [42].

Moreover, qualitative analysis of the distribution of correct responses across spatial quadrants revealed descriptive spatial biases in attentional deployment. Children with ADHD-H identified right-sided targets more accurately than left-sided ones, and a non-significant trend indicated better performance for upper compared to lower spatial positions. These findings suggest asymmetries in visuospatial exploration, consistent with reduced sustained and selective visual processing. Such rightward biases have been previously reported in visual or visuomotor tasks in ADHD-H and may reflect altered attentional allocation rather than focal neglect [46, 47].

A large body of literature has demonstrated the functional dominance of the right hemisphere for sustained and spatial attention, impulse control, and visuospatial integration. Furthermore, several studies have highlighted similarities between ADHD-H and attentional disturbance observed in patients with acquired right hemisphere dysfunction [48–50]. Within this framework, the spatial biases observed in the present study may be interpreted as reflecting altered efficiency within right-lateralized attentional networks, rather than focal structural impairment [48–51]. Repetitive response patterns were also observed in the ADHD-H group, consistent with reduced inhibitory control and cognitive flexibility, in line with core behavioral characteristics of ADHD [34]. Similar executive dysfunctions involving inhibition, set-shifting, and working memory have also been reported in preschool children with Global Developmental Delay, suggesting that early impairments in executive control may represent a transdiagnostic vulnerability factor across neurodevelopmental conditions [52].

From a neurofunctional perspective, these findings are compatible with models proposing that attentional deficits in ADHD arise from delayed or inefficient functioning within the frontoparietal attention network. Both the prefrontal and posterior parietal cortices play a crucial role in top-down modulation of visual attention and in maintaining goal-directed focus during complex visuospatial processing [53].

Structural and functional neuroimaging studies in individuals with ADHD have identified alterations in multiple brain regions, including the dorsolateral and ventrolateral prefrontal cortices and fronto-striatal circuits [54–57]. These findings provide a neurobiological framework within which attentional inefficiencies observed behaviorally may be understood, although direct neural inferences cannot be drawn from the present data.

In line with our hypothesis, children with ADHD-H performed significantly worse than controls on the RCPMs total score and on the two components requiring higher-order visuospatial integration and abstraction, such as gestalt reasoning and analogy processes. Conversely, the similarity component, involving simpler perceptual matching, did not differ significantly between groups. This dissociation suggests that basic perceptual processes are relatively preserved in ADHD-H, whereas impairments emerge when attentional control and integrative reasoning are required. This pattern reinforces the view that deficits in attentional control and visuospatial processing constrain performance on fluid intelligence tasks that depend on sustained manipulation and integration of visual information [58]. Accordingly, lower performance on nonverbal intelligence measures in ADHD-H should not be interpreted as evidence of reduced general intelligence, but rather as secondary to attentional dysregulation.

Consistent with this interpretation, no significant differences emerged between groups on the Verbal Abstraction Test. The comparable performance of children with ADHD-H and controls suggests that verbal reasoning and semantic abstraction abilities are relatively preserved. This dissociation between verbal and visuospatial modalities supports the view that cognitive deficits in ADHD-H are domain-specific rather than global, with greater vulnerability in tasks requiring visuospatial exploration and sustained attentional engagement.

Our results may have significant diagnostic and rehabilitative implications. First, they suggest that clinicians, educators, and intervention programs should be careful when interpreting low scores on fluid intelligence tasks in children with ADHD. Poor performance on visually mediated nonverbal tasks such as the RCPMs may reflect attentional instability rather than intellectual delay. Misinterpreting such scores as indicators of low general intelligence could lead to diagnostic inaccuracies and inappropriate educational strategies.

Second, the findings emphasize the need for targeted neurocognitive training aimed at improving visuospatial attention, impulse control, and strategic visual exploration. Such interventions could not only enhance attentional efficiency but also indirectly strengthen reasoning and learning capacities. Given the strong relationship between attention, executive functioning, and academic success, programs fostering attentional control could yield broad developmental benefits.