Attention Deficit/Hyperactivity Disorder (ADHD) is a neurodevelopmental disorder with altered levels of attention, and/or hyperactivity/impulsivity (1). Usually considered as a childhood disorder, around 40-60% of children with ADHD show symptoms during adulthood leading to a worldwide prevalence of 2.5% in adulthood (2–4). ADHD has been linked to dysregulation within the brain’s dopamine system, particularly in regions critical for reward processing and executive function, such as the prefrontal cortex and the striatum (5, 6). Altered dopamine transmission in individuals with ADHD is thought to contribute to impairments in motivation, attention, and impulse control (7). This dysfunction is further reflected in structural and functional alterations across key neural networks (8, 9). Functional connectivity studies have consistently shown disruptions in fronto-striatal and fronto-parietal circuits, which are essential for cognitive control (10). Furthermore, individuals with ADHD often exhibit reduced synchronization between the default mode network (DMN) and task-positive network (TPN), which impairs the brain’s capacity to switch between task-focused and resting states (11, 12). The hypoactivity of the dopamine system in ADHD may also lead to compensatory behaviors, such as hyperactivity or impulsivity, as the brain seeks stimulation to normalize dopamine levels (13). The neurotransmitter dopamine is synthesized through a process that depends on iron as a critical cofactor for the enzyme tyrosine hydroxylase (14). This enzyme catalyzes the rate-limiting step in dopamine production, meaning that adequate iron levels are essential for maintaining proper dopamine function. New neuroimaging advancements allow for precise brain iron quantification by applying quantitative susceptibility imaging (QSM) (15). QSM is particularly suited for measuring brain iron because iron is a paramagnetic substance, which alters the local magnetic field and induces changes in magnetic susceptibility detectable via MRI. These susceptibility changes can be quantified and mapped, providing a reliable estimate of iron concentrations in various brain regions. Importantly, QSM offers higher sensitivity and specificity compared to other imaging techniques, such as T2* or R2* relaxation mapping, allowing for more accurate inferences about brain iron distribution (15–17). In children with ADHD, only a handful of studies exist, investigate brain iron by applying various techniques (18). Preliminary evidence points to lower brain iron in children with ADHD, especially in the basal ganglia e.g., Putamen, Globus Pallidum, and Nucleus Caudate (19–23). On the other hand, some studies suggest that differences in brain iron levels between individuals with ADHD and controls are not altered and that variations in brain iron may instead be influenced by comorbid conditions, such as internalizing disorders (24–26). It is worth noting that most studies either did not examine associations with ADHD symptoms or failed to demonstrate significant findings, due to not surviving corrections for multiple comparisons. One study by Shvarzman was able to show a negative relationship between limbic striatal regions and symptom severity on Conners inattentive and hyperactive subscales, whereas Chen et al. reported a linear relationship between left anterior cingulum and symptom severity (19, 20).

As previously mentioned, while ADHD has its origins as a neurodevelopmental disorder, it persists throughout the lifespan, with significant alterations in dopaminergic pathways continuing to manifest in adulthood (3, 27). Recently, the first quantitative susceptibility mapping (QSM) data in adults with ADHD reported regionally increased magnetic susceptibility—interpreted as elevated iron load—in precentral, parietal, cingulate, and striatal territories, and linked higher precentral susceptibility to higher plasma neurofilament light (NfL) (28), suggesting possible neuroaxonal vulnerability and an association with dementia risk later in life. These findings suggest that atypical brain susceptibility is not restricted to pediatric ADHD and raise the possibility that altered susceptibility may have clinical relevance in adulthood. An open question, however, is how regional susceptibility in adults with ADHD relates to core ADHD symptomatology itself. Additionally, it remains unclear (i) whether group differences in susceptibility are detectable in regions supporting attentional selection, salience evaluation, and motivation, (ii) whether such differences are robust to individual variability in depressive symptom severity, which is common in adult ADHD and differed between patients and controls in the present sample, and (iii) whether regional susceptibility scales with current and retrospective ADHD symptom burden. The current study addresses these points by quantifying regional magnetic susceptibility in adults with ADHD and matched controls using QSM. We further examine whether susceptibility in these regions relates dimensionally to current ADHD symptom expression (Conners’ Adult ADHD Rating Scales) and retrospective childhood symptoms (Wender Utah Rating Scale). Based on pediatric evidence of altered striatal susceptibility in ADHD and emerging adult data indicating susceptibility differences in motor and associative cortex, we hypothesize that adults with ADHD would show regionally specific susceptibility alterations relative to controls, and that susceptibility in these regions would covary with ADHD symptom burden.

This study examined regional magnetic susceptibility in adults with ADHD relative to neurotypical individuals using quantitative susceptibility mapping (QSM) and related these measures both to diagnostic group status and to dimensional symptom measures. In contrast to our initial expectations, no ROI survived correction for multiple comparisons, and overall group differences were very subtle. This suggests that global brain iron homeostasis may be largely intact in our adult ADHD sample. Nevertheless, we did observe a few nominal effects that merit discussion. Adults with ADHD showed lower brain iron susceptibility in several cortical regions, notably ventral temporal and limbic areas, relative to controls (e.g., the fusiform gyrus and posterior cingulate cortex). These differences were small and did not survive false discovery rate correction, but their regional pattern is intriguing. The absence of robust case-control differences in brain iron is generally in line with recent findings in younger cohorts. A QSM study of adolescents reported no significant differences in subcortical iron between ADHD and typically developing groups (24), suggesting that alterations of brain iron deficit may not be a consistent feature in adult ADHD. Earlier MRI investigations using iron-sensitive methods did hint at iron abnormalities in ADHD, but those effects appeared tied to developmental stage and treatment exposure. For example, medication-naïve children with ADHD have been found to have significantly lower iron indices in the putamen, caudate, and thalamus compared to controls, whereas children on psychostimulants showed iron levels indistinguishable from healthy (18). This suggests that brain iron differences in ADHD may be most apparent early in the disorder or in unmedicated individuals, and that maturation and/or treatment over time may normalize these differences. By adulthood, many individuals with ADHD could have “caught up” in terms of iron deposition, consistent with the broader pattern of neuroanatomical development in ADHD, where delays in childhood tend to diminish later on (44–46). Our null group findings support the idea that widespread brain iron deficits or excesses are not present in the average adult with well-managed ADHD. In other words, if brain iron dysregulation contributes to ADHD symptoms in youth (perhaps via reduced cofactor availability for dopamine), those effects may be transient or compensated for by adulthood. It is important to note, however, that the literature on adult ADHD brain iron is still sparse and not entirely uniform. Contrary to our results, one recent study reported significantly higher QSM-measured iron in adults with ADHD relative to (28). In that study, the ADHD group showed elevated susceptibility in multiple regions, with the largest effect in the precentral cortex, and a corresponding increase in a blood neuroaxonal injury marker. The authors interpreted this as evidence of excess iron accumulation in ADHD that could render the brain more vulnerable to neurodegenerative processes. In our sample, we did not observe comparable effects: the precentral gyrus and other regions emphasized in that report (e.g., anterior cingulate, precuneus, and basal ganglia) did not exhibit statistically significant case-control differences. In addition to methodological differences, stimulant medication status may have contributed to the discrepant pattern of findings. In the present study, participants were instructed to refrain from ADHD medication (n=5/25) for at least 24 h before MRI acquisition to minimize potential pharmacological effects. In contrast, Berberat et al. reported a naturalistic medication regimen in the ADHD group, with ~59% receiving medication not altered in the context of study participation (28). Although direct evidence in adults is limited, pediatric MRI work suggests that psychostimulant exposure is associated with a relative “normalization” of brain iron indices: medication-naïve youth with ADHD show reduced striatal/thalamic iron markers compared with controls, whereas chronically medicated youth tend to resemble control levels, and longer treatment duration has been linked to a greater degree of normalization (18, 24, 47, 48). Taken together, differences in medication exposure and the clinical characteristics that covary with medication status could plausibly shift regional susceptibility estimates and thereby contribute to the divergence between studies.

In terms of ADHD symptoms, we observed several brain-behavior correlations that, while not significant after controlling for multiple comparisons, suggest potentially meaningful inter-individual variation in susceptibility measures. In particular, susceptibility in the fusiform gyrus showed nominal negative associations with ADHD symptom severity across both current and retrospective indices, such that higher symptom burden tended to relate to lower susceptibility in ventral temporal cortex. Nominal associations were also observed in posterior midline and parietal association regions (precuneus and posterior cingulate cortex, and, for impulsivity, additionally superior parietal cortex), consistent with the involvement of attentional control and default-mode networks in ADHD (49–51). These findings imply that even in the absence of large case-control differences, individual differences in brain iron may track with the severity and history of ADHD. One interpretation is that lower iron in certain cortical regions could be a marker of delayed or altered neurodevelopment associated with more severe ADHD. For instance, the fusiform is involved in higher-order visual processing and attention (52, 53) and prolonged developmental lag or reduced metabolic activity in these regions might manifest as both reduced iron incorporation and greater attentional symptoms. This aligns with the concept that ADHD’s neurobiology may reflect a timing difference in brain maturation rather than a permanent deficit (45, 46). Individuals with the most pronounced or persistent ADHD symptoms might show subtle “fingerprints” of that delayed maturation in measures like QSM.

In exploratory analyses of depressive symptoms, higher BDI scores were associated with lower susceptibility in two default-mode network hub regions-the posterior cingulate cortex and precuneus. In contrast, a positive association emerged in the bilateral occipital cortex. Importantly, none of these correlations survived FDR correction and thus should be interpreted as hypothesis-generating. The direction of the PCC/precuneus associations is nonetheless biologically plausible given (i) the central role of PCC/precuneus within the DMN in depression-related self-referential processing and rumination, and (ii) convergent MRI-QSM evidence in depression showing susceptibility alterations in cingulate regions, including the PCC, in relation to illness course and progression (54–56). From a mechanistic perspective, depression has been linked to dysregulation in iron homeostasis, i.e., iron-related oxidative stress and neuroinflammatory pathways are plausible modulators of regional susceptibility, potentially contributing to subtle DMN-related tissue differences (57, 58).

After adjusting for sex, no age–QSM associations survived correction for multiple comparisons; however, at the uncorrected level, age was positively associated with susceptibility in posterior cingulate and frontotemporal regions. This pattern is consistent with QSM evidence that brain susceptibility tends to increase across adulthood, with particularly robust age effects reported in deep gray matter and more region-specific effects in cortex (59). Although cognitive performance was not the primary focus here, prior work suggests that higher cortical iron/susceptibility can be related to age-associated reductions in fluid cognition and executive functioning, motivating future lifespan studies that jointly model susceptibility and cognition in ADHD (60, 61).

Taken together, our findings paint a nuanced picture of brain iron in adult ADHD. On one hand, the overall null result for case–control differences indicates that altered brain iron levels may not be a defining feature of ADHD in adulthood. The dopaminergic dysfunction in ADHD is more likely attributable to synaptic and receptor-level mechanisms than to a lack of the iron cofactor in dopamine-producing regions. On the other hand, the pattern of nominal effects and correlations we observed hints that brain iron metrics can still yield insight into ADHD’s pathophysiology when examined in detail. The regional specificity of the nominal group differences suggests that subtle microstructural or metabolic differences persist in these circuits. Notably, the posterior cingulate and orbitofrontal cortex are nodes of the default mode and reward networks, which have been reported to be functionally atypical in ADHD (62, 63). If those network differences have a developmental basis, they might leave an imprint on physical tissue properties such as iron content or myelination. Thus, while small, the susceptibility differences in these regions align with the broader understanding of ADHD as involving fronto-limbic and fronto-striatal dysregulation. The fact that these differences did not survive correction emphasizes that they are subtle; yet, they were consistent in direction across multiple analytical models, lending them some credibility. Furthermore, the clinically relevant coupling between susceptibility and symptom burden that we found supports the biological plausibility of our measures. It suggests that QSM, beyond group comparisons, may capture meaningful variation related to ADHD’s clinical heterogeneity.

Brain iron estimated has been proposed as a candidate biomarker for ADHD because iron is tightly coupled to dopaminergic function and because MRI-based approaches provide an in vivo, regionally specific readout of tissue susceptibility that may be more sensitive to neurobiological variation than peripheral iron indices (18, 23). In pediatric ADHD, converging evidence suggests that altered brain iron may be detectable and potentially clinically informative, although larger longitudinal studies are required, and the specificity relative to other neurodevelopmental disorders remains unresolved. At the same time, several considerations temper the plausibility of QSM-derived brain iron as a diagnostic biomarker in adults with ADHD. First, adult brain iron is strongly age-dependent and influenced by cumulative exposures (including long-term stimulant treatment histories, lifestyle factors, and cardiometabolic or inflammatory processes), complicating the definition of stable, diagnostically useful thresholds across heterogeneous adult cohorts. Second, in the present data, ADHD-control differences were limited to nominal effects that did not survive multiplicity correction and were sensitive to covariate specification (e.g., adjustment for depressive symptoms), arguing against a robust case-control signature suitable for diagnostic classification. Third, the regional pattern observed here diverges from the canonical pediatric basal-ganglia-centric biomarker narrative, raising the possibility that adult susceptibility differences reflect downstream or compensatory processes, developmental “catch-up,” or comorbidity-related mechanisms rather than a core, disorder-specific marker. Another important interpretive consideration is that associations between QSM-derived susceptibility and lifetime ADHD burden may plausibly reflect bidirectional processes. Brain iron estimates show pronounced age dependence across adulthood and may index multiple microstructural contributors beyond iron, complicating a static trait interpretation in adult cohorts (64). From a consequence perspective, ADHD-related behavioral and clinical exposures (e.g., chronic sleep disruption, stress, and metabolic/inflammatory comorbidity) could influence iron homeostasis and redox-related biology over time. Converging literatures link sleep-wake disturbances and inflammatory-metabolic states to iron regulation and iron-related pathways (65). Conversely, inter-individual differences in brain iron-related susceptibility may contribute to symptom persistence by modulating dopamine-relevant neurobiology and network efficiency, consistent with the broader rationale for investigating brain iron as a candidate biomarker in ADHD, particularly in pediatric work, while acknowledging that adult heterogeneity and cumulative exposures may weaken diagnostic specificity.

Overall, this study provides no robust evidence for altered brain iron susceptibility in adults with ADHD. No case-control differences survived multiple comparison correction. At the nominal (uncorrected) level, ADHD was associated with lower susceptibility in a small set of regions, most consistently the right fusiform and left posterior cingulate cortex. While these patterns are neurobiologically interpretable (e.g., higher-order ventral visual/temporo-limbic and DMN-related hubs), they should be treated as exploratory. Collectively, the findings support a cautious interpretation: if QSM differences in adult ADHD exist, they are likely small-to-moderate and regionally nuanced, and future work will require larger, medication-stratified, developmentally informed samples to clarify whether subtle iron-related signatures track symptom dimensions, treatment exposure, and neurodevelopmental timing.