ADDitude for Professionals

Neuroimaging and ADHD: Findings, Limitations, and Promise

Neuroimaging promises to advance our understanding of ADHD’s biological underpinnings and to ultimately facilitate diagnosis and screening, improving treatment outcomes for children and adults. While important studies exist, the field has yet to translate available research and insights to the clinical realm. For ADHD neuroimaging to reach its potential, researchers must address these challenges and finding new areas of exploration.

By Barbara Franke, Ph.D., Hugh Garavan, Ph.D., Tor Wager, Ph.D. Verified Updated on June 15, 2022
Click to Add Comments
Magnetic resonance imaging

The buzz around neuroimaging — and specifically its promise as a tool for understanding attention deficit hyperactivity disorder (ADHD or ADD) — has grown louder in recent years. Researchers are working now to determine how neuroimaging, including functional magnetic resonance imaging (fMRI) and other imaging techniques, might reveal insights about the brain structures and measures potentially implicated in ADHD. In essence, neuroimaging facilitates the collection of biological measurements of the brain, aiming to increase our understanding of the biological underpinnings of ADHD and potentially facilitate the application of findings in clinical settings to aid in diagnosis and treatment.

In recent years, neuroimaging studies for ADHD have yielded some significant developments and promising directions for further exploration. At the same time, efforts are underway to resolve a range of challenges, limitations, and barriers to robust analysis and meaningful applications.

Neuroimaging and ADHD: Developments and Challenges

By identifying biological measures for ADHD, researchers can offer substantial and nuanced new ways of characterizing this heterogeneous disorder, which appears to be rooted in genetic, environmental, and neural factors. Brain measures can be used to develop key biomarkers, including:

  • Diagnostic biomarkers, which link a brain structural measure, activity pattern, or conductivity to a particular diagnostic category.
  • Pharmacodynamic/response biomarkers, which reveal if treatment strategies are impacting the intended brain mechanisms, with a potential impact on symptoms and disease severity.
  • Prognostic biomarkers, which predict the development of a phenotype or a comorbid disorder in the future.

Ultimately, scientists hope to use these biomarkers to aid in areas like early detection and stratification, and to uncover a basis for ADHD heterogeneity that may improve diagnostic and treatment approaches.

[Read: Face It — People with ADHD Are Wired Differently]

Important advancements and findings in ADHD neuroimaging have emerged in recent years. Neuroimaging studies show structural distinctions in several brain regions, especially in children with ADHD. A 2015 review1, for example, summarized brain mechanisms across multiple modalities and the differences between controls and individuals with ADHD.

However, the findings and literature on ADHD neuroimaging still have multiple limitations, including but not limited to:

  • Small sample size in a vast majority of studies, possibly resulting in inflated effect sizes of observed brain alterations and lack of detection of other brain alterations.
  • An overrepresentation of children with ADHD, leaving adolescent and adult ADHD understudied.
  • A traditional focus on region-by-region brain mapping rather than looking at the whole brain, and how parts of the brain function together. This leads to problems like irreproducible results, low reliability, and low power with small sample-size studies, among other issues.

Sample Sizes and Small Effects

Large sample sizes are needed for robust analysis in neuroimaging. In its search for brain correlates in ADHD, the neuroimaging field might therefore benefit from resetting expectations on findings, especially on just how large we assume effect sizes must be. Statistically, samples with a smaller number of participants result in substantial variation. Most studies in the neuroimaging field, though, tend to include 100 participants or fewer. The result of this is inflated effect sizes in the literature, which also suffer from publication bias, where only positive findings tend to be published.

[Read: Could High-Tech Brain Scans Help Diagnose ADHD?]

The rise of big data in neuroimaging is helping to address these issues. Take the ENIGMA Consortium, founded in 2009, which created an international network of brain imaging data for researchers across multiple disciplines to access. The data collected as part of the ENIGMA ADHD Working Group paved the way for a 2017 mega-analysis of subcortical volumes (regions like the amygdala, thalamus, etc.), hippocampus and intracranial volume (a measure of total brain volume) in ADHD, with the aim of addressing weaknesses in prior imaging studies.

With more than 1,700 participants with ADHD and 1,500 participants without ADHD, ranging from ages 4 to 63 years, the study – the largest in ADHD at the time – found slightly lower volume in most of the brain’s subcortical regions among individuals with ADHD, compared to controls2. Further analysis showed that these measures were largely present in children, with effects attenuated in adults. The study also showed that sample size remains an issue in imaging studies for ADHD.

Predictive Modeling and Biomarkers

The neuroimaging field is steadily moving closer to identifying predictive features and biomarkers for ADHD. A 2019 ENIGMA-ADHD study3 on cortical features (i.e. surface area of brain regions and brain thickness) with over 2300 participants with ADHD and over 2000 participants without ADHD found that children with ADHD showed smaller structures in several parts of the brain — namely the frontal and orbitofrontal cortex, the cingulate cortex, and the temporal cortex — compared with controls. While the study included adolescent and adult participants, no significant effects were seen in these groups. In fact, the younger the children, the larger the effect on the brain structure. The study also revealed another important finding: acute ADHD symptoms and attention problems, as assessed in children from the general population, are associated with significantly smaller brain surface area regions in the same regions as found altered in the cases.

An earlier neuroimaging study involving ADHD symptomatology and cognitive tests saw similar findings. Using a longitudinal European sample of about 2,000 children, this 2017 study4 found that parent and youth ratings of ADHD symptoms were negatively associated with gray matter volume in the ventromedial prefrontal cortex (vmPFC), which supports existing literature that links this region with ADHD symptoms. Moreover, the study found that these brain effects predict symptomatology five years later, possibly indicating that the vmPFC is a biomarker for ADHD.

Building off the aforementioned 2017 and 2019 ENIGMA-ADHD findings, a recent follow-up study5 explored whether that data could sufficiently predict ADHD case status in children and/or adults. After applying deep learning algorithms, the study found that there is, in fact, predictive value to the data for both. Furthermore, the deep learning model, when trained on adult ADHD data, could actually predict the childhood ADHD data. This shows that, despite no significant effects in this group, there is information in the adult brain that links it to ADHD. The predictions, while insufficient for clinical use, are a critical step for future modeling.

Neuroimaging and ADHD: Promising Directions

Given current limitations and available data, in what direction should ADHD neuroimaging head? How can researchers improve on studies and begin to find stronger, more robust associations between brain measures and ADHD? Attending to the heterogeneity of ADHD, e.g. through subgrouping, may be one viable pathway.

Heterogeneity

ADHD is highly heterogeneous, varying in presentation from individual to individual. And yet the vast majority of neuroimaging studies assume a clear distinction between patients and controls. Grouping ADHD individuals together — regardless of subtypes and individual differences — may badly hurt our ability to find consistent, reliable, and robust measures correlated to symptoms.

Indeed, a recent study that applied a novel normative model to participants with ADHD found that the group deviated from the model overall, but that there was limited overlap at the individual level, indicating that heterogeneity in brain alterations is strong between adult individuals with ADHD6.

Moving away from the “average ADHD patient” approach could provide the neuroimaging field with more useful data. While not many studies concentrate on individual patients, yet, subgrouping efforts have been going on in the field.

Subgrouping Efforts

A new study using ENIGMA-ADHD Working Group data was able to find that subgrouping algorithms may reveal more robust effect sizes in studies of structural brain imaging data of ADHD7. The study analyzed subcortical volume data from boys with and without ADHD subdivided into three distinct areas (factors): the basal ganglia, the limbic system, and thalamus. Based on these factors, participants could be separated into four distinct “communities” or subgroups. The results of the study showed that the effect sizes of case-control differences were larger within individual communities than they were in the total sample.

Continuing to explore and organize according to ADHD heterogeneity, including the degree to which inter-individual differences exist, may provide important insights to inform future neuroimaging research.

Neuroimaging for ADHD: Next Steps

SUPPORT ADDITUDE
Thank you for reading ADDitude. To support our mission of providing ADHD education and support, please consider subscribing. Your readership and support help make our content and outreach possible. Thank you.

View Article Sources

1 Faraone, S., Asherson, P., Banaschewski, T. et al. Attention-deficit/hyperactivity disorder. Nat Rev Dis Primers 1, 15020 (2015). https://doi.org/10.1038/nrdp.2015.20

2 Hoogman, M., Bralten, J., Hibar, D. P., Mennes, M., Zwiers, M. P., Schweren, L., van Hulzen, K., Medland, S. E., Shumskaya, E., Jahanshad, N., Zeeuw, P., Szekely, E., Sudre, G., Wolfers, T., Onnink, A., Dammers, J. T., Mostert, J. C., Vives-Gilabert, Y., Kohls, G., Oberwelland, E., … Franke, B. (2017). Subcortical brain volume differences in participants with attention deficit hyperactivity disorder in children and adults: a cross-sectional mega-analysis. The lancet. Psychiatry, 4(4), 310–319. https://doi.org/10.1016/S2215-0366(17)30049-4

3 Hoogman, M., Muetzel, R., et al (2019, April 24). Brain imaging of the cortex in ADHD: A coordinated analysis of large-scale clinical and Population-based samples. AM J Psychiatry. https://doi.org/10.1176/appi.ajp.2019.18091033

4 Albaugh, M. D., et al (2017). Inattention and Reaction Time Variability Are Linked to Ventromedial Prefrontal Volume in Adolescents. Biological Psychiatry, 82(9), 660–668. https://doi.org/10.1016/j.biopsych.2017.01.003

5 Zhang-James, Y., Helminen, E.C., Liu, J. et al. Evidence for similar structural brain anomalies in youth and adult attention-deficit/hyperactivity disorder: a machine learning analysis. Transl Psychiatry 11, 82 (2021). https://doi.org/10.1038/s41398-021-01201-4

6 Wolfers, T., Beckmann, C. F., Hoogman, M., Buitelaar, J. K., Franke, B., & Marquand, A. F. (2020). Individual differences v. the average patient: mapping the heterogeneity in ADHD using normative models. Psychological medicine, 50(2), 314–323. https://doi.org/10.1017/S0033291719000084

7 Li, T. et al. (2021). Characterizing neuroanatomic heterogeneity in people with and without ADHD based on subcortical brain volumes. J Child Psychol Psychiatry, in press. bioXiv doi: https://doi.org/10.1101/868414