Attention Deficit Hyperactivity Disorder (ADHD) is a common neurodevelopmental disorder characterized by symptoms of inattention and hyperactivity-impulsivity, affecting approximately 5%-7% of school-aged children globally (1, 2). Regional variations exist, with approximately 10% prevalence in the United States (3) and 6.4% in China (4). Despite clear diagnostic criteria (DSM-5, ICD-11), ADHD treatment depends on parent-reported symptoms, and their remission is key to better function and quality of life (5). However, significant symptom heterogeneity complicates diagnosis and treatment, necessitating investigations into underlying influencing factors.

Both the executive function hypothesis and the cognitive processing theory posit that cognitive functional deficits are considered the core impairments of ADHD, with cognitive deficits potentially serving as endophenotypes of ADHD. A recent systematic review identified consistent executive function (EF) impairments in children with neurodevelopmental disorders, particularly in attention, response inhibition, working memory, and planning (6). These EF deficits correlate with ADHD core symptoms. Emerging evidence suggests symptom dimension–specific neural substrates: inattention is linked to deficits in attentional control, whereas hyperactivity-impulsivity is associated with impaired response inhibition (7). Complementing the EF framework, the cognitive processing theory proposes that disruptions in distinct cognitive operations—such as planning, attention, and successive processing—underlie ADHD pathology. Empirical studies using cognitive processing assessments have confirmed these deficits in children with ADHD (8, 9). However, the direct relationships between cognitive processing components and specific ADHD symptom dimensions remain less thoroughly explored than EF-based models. Clarifying these associations may deepen our understanding of the distinct cognitive-neural mechanisms underlying inattention and hyperactivity-impulsivity.

Comorbidities, notably behavioral issues and sleep disorders, frequently co-occur with ADHD, complicating clinical presentations and potentially exacerbating ADHD symptom severity. Approximately 50%-90% of children with ADHD exhibit at least one comorbid condition, such as learning disabilities, oppositional defiant disorder (ODD), or anxiety disorders (4, 10, 11). Additionally, prevalent sleep disturbances like difficulty initiating and maintaining sleep, excessive daytime somnolence, and sleep-wake transition disorders significantly correlate with worsened ADHD symptoms (12–15). These issues highlight the importance of comprehensively understanding how cognitive, behavioral, and sleep-related factors collectively influence ADHD symptom profiles.

However, critical gaps remain regarding the interactive roles of demographic characteristics, cognitive functioning, sleep disturbances (when considered as specific phenotypes rather than a unitary construct), and behavioral comorbidities in shaping ADHD symptomatology. In particular, the differential effects of discrete sleep phenotypes (e.g., disorders of initiating and maintaining sleep [DIMS], disorders of excessive somnolence [DOES], and sleep–wake transition disorders [SWTD]) and the modulatory role of emotional profiles—such as the reported inverse association between anxiety and hyperactivity-impulsivity—have been understudied. Addressing these gaps is pivotal for advancing precision diagnostics and developing tailored interventions. Therefore, in the present study we adopt a multidimensional framework that integrates demographic, cognitive (DN: CAS), sleep (SDSC subscales), and behavioral (Conners PSQ) measures to examine predictors of inattention and hyperactivity-impulsivity, with particular attention to specific sleep phenotypes and the low-anxiety/high-impulsivity relationship. This focused, integrative approach aims to clarify symptom-specific pathways and inform more individualized assessment and intervention strategies.

This study identified distinct predictors for the two core dimensions of ADHD. Hyperactivity-impulsivity symptoms decreased with age and were more pronounced in boys, whereas inattention symptoms remained stable and showed no gender difference. Inattention was primarily associated with attentional deficits and learning problems, while hyperactivity-impulsivity was more strongly linked to oppositional behaviors, sleep-wake transition disturbances, learning problems, and low anxiety levels, suggesting symptom-specific developmental and behavioral pathways. Importantly, the novelty of our work lies in its multidimensional, phenotype-focused approach: we simultaneously assessed demographic, cognitive, behavioral, and discrete sleep phenotypes, and explicitly examined the counterintuitive association between lower anxiety and greater hyperactivity-impulsivity. This focused integration provides more specific evidence on symptom-dimension–specific correlates than studies that treat sleep or behavior as unitary constructs.

The developmental trends and gender differences observed in this study are consistent with previous findings. Hyperactivity-impulsivity symptoms declined with age, while inattention remained stable—mirroring longitudinal results from Holbrook et al. (22) and reinforcing that inattention may persist regardless of ADHD subtype (23). Boys exhibited higher levels of hyperactivity-impulsivity, with no significant sex difference in inattention, in line with meta-analytic evidence (24, 25). These patterns highlight the risk of underdiagnosing girls with inattentive presentations, as noted by Carucci et al. (26), and support calls for gender-sensitive screening focused on subtle cognitive disengagement. Beyond demographic patterns, our study extends prior work by identifying symptom-specific predictors. Attentional deficits assessed by DN: CAS were selectively associated with inattention, refining earlier general cognitive findings (8, 9). We also found that SWTD independently predicted hyperactivity-impulsivity, providing a more nuanced view of sleep phenotype effects (12, 13). Finally, our finding that low anxiety was linked to greater impulsivity supports previous suggestions that anxiety may suppress overt hyperactive behaviors (27), underscoring its modulatory role in ADHD symptom expression.

The findings that inattention symptoms were significantly associated with attentional dysfunction and learning problems, while hyperactivity-impulsivity symptoms were primarily predicted by oppositional behavior, specific sleep disturbances, and low anxiety, suggest distinct neuropsychological pathways underlying the two symptom dimensions of ADHD. Deficits in attentional function, as assessed by the DN: CAS, reflect impairments in selectively maintaining focus, inhibiting irrelevant stimuli, and allocating processing resources under cognitive load. The Attention subtest includes Expressive Attention, Number Detection, and Receptive Attention, collectively targeting interference control and vigilance (8). These deficits directly impact classroom learning and explain the strong association with academic difficulties. Prior studies have demonstrated that children with ADHD show persistent impairments in attention and planning, which disrupt information processing efficiency (9). In our findings, attentional performance predicted inattention but not hyperactivity, indicating that the two symptom clusters may involve non-overlapping cognitive substrates. Sleep disturbances also displayed symptom-specific associations. Although DIMS and DOES initially correlated with inattention, these effects diminished after accounting for behavioral comorbidities, suggesting possible mediation. In contrast, SWTD—which include phenomena such as sleep talking, nightmares, and myoclonic jerks—remained a significant predictor of hyperactivity-impulsivity even after full adjustment. SWTD reflects instability in arousal regulation during sleep, which may extend into the daytime as behavioral dysregulation and poor impulse control (13). Disrupted sleep continuity may impair prefrontal cortical functioning and lead to heightened emotional reactivity and motor overactivity, especially in young children with neurodevelopmental vulnerabilities.

Conduct problems emerged as the strongest behavioral predictor of hyperactivity-impulsivity. This is consistent with prior research indicating symptom overlap between ADHD-HI and ODD, where shared behavioral descriptors (e.g., “interrupts,” “argues,” “defies”) may reflect underlying deficits in emotional regulation and impulse inhibition (25). The significant independent contribution of ODD symptoms supports the hypothesis that HI and conduct problems may co-arise from prefrontal–limbic dysfunction and reinforce one another through negative behavioral-emotional feedback loops. An additional novel finding is the inverse relationship between anxiety and HI symptoms. Previous research suggests that anxiety may inhibit externalizing behaviors through heightened sensitivity to criticism or punishment (27). In contrast, children with low anxiety may exhibit under-responsiveness to threat cues and increased risk-taking. Animal models such as SHRSP rats demonstrate that reduced anxiety correlates with impulsive behavior, linked to alterations in hippocampal plasticity and serotonergic pathways. These mechanisms may help explain why low-anxiety children in our study displayed elevated HI symptoms, suggesting a potential “low-anxiety/high-impulsivity” ADHD subtype with distinctive neurobiological underpinnings and potentially lower responsiveness to standard psychostimulant treatments. Children with ADHD show deficits in multiple neural circuits, including frontostriatal dopaminergic loops, corticostriatothalamic pathways, frontocerebellar networks, and reward circuits. These abnormalities manifest as impairments in attention, cognitive control, timing, and reward processing. Collectively, these findings support a multi-pathway model in which distinct neural dysfunctions contribute to the heterogeneous clinical presentation of ADHD.

The dimensional dissociation of ADHD symptoms observed in this study suggests the need for tailored clinical strategies. Inattention, associated with cognitive and learning deficits, may benefit from cognitive training and academic support, while hyperactivity-impulsivity, linked to oppositional behaviors and sleep-wake transition disturbances, may require behavioral management, emotion regulation, and sleep-focused interventions. The inverse relationship between anxiety and impulsivity points to a potential low-anxiety, high-impulsivity subtype that may respond poorly to conventional stimulants. Moreover, the underrecognition of inattentive symptoms in girls highlights the need for gender-sensitive screening approaches to improve diagnostic accuracy and early intervention.

This study has several limitations. First, as a cross-sectional study, causal or directional inferences cannot be drawn; longitudinal designs are needed to clarify temporal relationships among cognition, sleep phenotypes, behavior, and symptom dimensions. Second, all symptom and comorbidity measures were parent−reported, which may introduce recall or reporter bias and could underdetect internalizing symptoms such as anxiety; future studies should incorporate multi−informant assessments (parent, teacher, and child self−report). Third, sleep assessment relied on the parent−rated SDSC, which—while practical for large samples—cannot capture physiological sleep parameters, prospective sleep–wake patterns, or circadian preference; we did not collect sleep diaries, actigraphy, or morningness–eveningness measures in this cohort. Fourth, we did not assess restless sleep or Restless Legs Syndrome (RLS) symptoms, which have been implicated in ADHD and may share dopaminergic mechanisms; targeted RLS screening should be included in future work. Fifth, although we statistically controlled for full−scale IQ in multivariable analyses to reduce confounding by intellectual functioning, interactions among IQ, attention, and learning problems are complex and warrant more detailed investigation, including stratified or mediation analyses. Taken together, future studies should use longitudinal designs with multi−informant ratings, objective sleep monitoring (e.g., actigraphy or polysomnography), prospective sleep diaries, circadian preference scales, and RLS assessment to better delineate causal pathways and to validate the symptom−dimension–specific associations reported here.

This study highlights the multidimensional nature of ADHD and the distinct predictors associated with its core symptom dimensions. Inattention and hyperactivity-impulsivity exhibit divergent developmental patterns and neurobehavioral correlates, underscoring the need for symptom-specific assessment and intervention strategies. Targeting cognitive deficits, behavioral comorbidities, sleep disturbances, and anxiety profiles may enhance individualized treatment planning. Future longitudinal and multi-informant research is warranted to clarify causal mechanisms and optimize precision care in ADHD.