Abstract
Research over the last decade has progressively found lack of support for the use of the Theta to Beta Ratio (TBR) as a diagnostic tool for ADHD. Archival samples drawn from the NewMind database support these findings and indicate the picture is even more complex with respect to EEG amplitude. Samples drawn from the database support findings that elevated theta is not correlated with attention problems but rather impulsivity and attention difficulties are often more related more to alpha slowing. Further analysis suggests that excess frontal delta is also correlated with attention difficulties.
Introduction
For two decades clinicians have been using TBR as an indicator confirming the presence of ADHD while the practice has been considered extremely controversial among researchers. The idea had its genesis when Joel Lubar (1991) proposed that the Theta to Beta Ratio (TBR) might be used as a diagnostic indicator of ADHD. Over that decade a dozen or so studies supported that hypothesis (Boxum et al., 2024) and it appeared to be very promising diagnostic instrument. Clarke et al. (1998) and Monastra et al. (1999) conducted key studies that found it to be highly accurate. Monastra et al. (1999) used a large sample of subjects (n = 482) and found TBR had a sensitivity of 86% and specificity of 98% when used to identify individuals with ADHD.
Snyder et al. (2008) did a study that further supported these findings resulting in the marketing and eventual FDA approval of the Neuropsychiatric EEG-Based ADHD Assessment AID (NEBA) in 2013 by NEBA Health in Augusta, GA. By this point in 2008 the TBR was commonly used by NFB practitioners and widely taught in workshops. The FDA approval appeared to bless the TBR as truly reliable and sound diagnostic indicator of ADHD. Yet this assumption ignored another developing thread of research in which the findings questioned the validity of that conclusion. Other researchers were obtaining results that forced them to draw other conclusions about ADHD that did not match up with previous research.
Arns et al. (2008) observed that many of their ADHD subjects had alpha slowing that was contributing to the higher theta in many cases. They suggested that researchers were confusing two sub-types of ADHD, those with true high theta and those with high amplitude alpha slowing which brought it into the theta range that was then detected by theta filters. They also noted that the slowed alpha may be related to sleep problems and subjects who displayed this pattern were non-responsive to medication. Lansbergen et al. (2011) supported this conclusion and felt the alpha slowing was actually more representative of ADHD. Not long after, Arns et al. (2013) published a meta-analysis noting other issues that interfered with replication and reliability in past research. Some studies used eyes closed and others used eyes open for the TBR ratio. Most used a two minute baseline but others varied from this standard. In addition authors varied on the location at which they measured TBR, with Cz and Pz being most popular while some even used FZ. Out of 70 studies between 1983 and 2011 they found only 9 studies that they felt could qualify for their analysis. Their conclusion was that TBR could not reliably distinguish between healthy neurotypical subjects and ADHD subjects. They felt their analysis further supported their hypothesis that slowed alpha was related to sleep issues.
This should have been a red flag for the neurofeedback community but most were only vaguely aware of this line of research and focused on the attention NEBA produced in 2013 as a diagnostic device. Unfortunately, two years after NEBA was approved for marketing by the FDA, Snyder et al (2015) failed to replicate their previous findings. Arns et al. and other researchers, including Barry Sterman, published an editorial in The Journal of Child Psychology and Psychiatry explaining the limits of the TBR and that its use is not included in the DSM-5 or best practice guidelines due to insufficient empirical support for its diagnostic validity. The FDA approval, they argued, does not equate device approval with best clinical practices or empirical support and that a significant gap existed between regulatory approval and clinical utility. Even today many practitioners continue to use TBR for evaluation without any knowledge of what the actual research indicates.
Kerson et al. (2019) and van Dijk et al., 2020 published articles highlighting the signal processing difference in softwares used in various studies that contributed substantially to confounds that may have led to false conclusions. Different softwares use different filter ranges for theta and beta. There is also the issue of large changes in theta amplitude in childhood as myelination progresses that are not taken into account in the fixed TBR. In addition without a hi beta (21-30Hz) inhibit, real time auto-artifact, or EMG monitoring the ratio has a high probability of being confounded by scalp EMG, which is ubiquitous among individuals with a disorder (Paluch, 2017; Soutar, 2023).
Clarke et al. (2020) published an article highlighting the heterogeneity of ADHD indicating EEG sub-clusters within the diagnosis. Mohagheghi et al. (2017) did a carefully designed RCT study comparing SMR training to high frequency alpha training and found subjects in the alpha training group improved more on attention. This further confirmed the hypothesis that alpha played more of a role in inattention than theta. Boxum et al. (2024) published an analysis confirming all these findings and argued that peak alpha was more an indicator of general arousal problems associated with ADHD than TBR and stating that the use of TBR ratio for diagnosis was definitely not recommended and that it lacked sensitivity and specificity. However, he also echoed Arns et al. and others by acknowledging that the distinction between real frontal theta and slowed frontal alpha was useful in distinguishing who might respond to medication and that the frontal theta responded better to theta inhibit training than the slowed alpha population.
Data Collection & Subjects
In consideration of the foregoing, it is of interest what light archival data might shed on these findings. Using the NewMind database we extracted a large sample of subjects for analysis. The data was taken from FZ because past research had found the TBR ratio valid at both locations and because FZ theta was more representative than Cz of frontal slowing in the eyes closed condition. The data was collected using the NewMind 20 channel amplifiers with auto-artifacting. The default baseline setting on NewMind software is two minutes, which the majority of clinicians use. The extreme outliers were eliminated form an n = 2393 composed of adults age 18 years and older to avoid the wide variation in theta that occurs in younger subjects and included both males and females.
The CEC had been used to assess all subjects. The attention dimension of the assessment tool is cross validated with the TOVA (Soutar, 2014). Subjects were divided into two groups, those with very low scores of 0-13 comprising the lower third of those assessed (least attention problems) and those with the highest third (most attention problems) with scores of 20 – 33.
Arns et al. 2013 suggested that the slow alpha sub-types in their sample may be related to sleep issues. Sleep issues are also ubiquitous in TBI as well as attention deficits. One common indicator of recovery from TBI is sleep improvement (Duclos et al., 2016). In light of the relationship between frontal delta and attention as well as cognitive function, it was decided that frontal delta should also be examined. Again, in this analysis, we compared the lowest scoring third (least attention problems) with the highest scoring third (most attention problems) on the CEC with delta baseline amplitude.
Results
All comparison are based upon qEEG eyes closed measures. Comparison of theta amplitudes in the lowest scoring third of the group was compared with theta in the highest scoring group using a t-test (Two-Sample Test Assuming Unequal Variances). The mean of the low scores (good attention) was 8.82 and the mean of the high scores (poor attention) was 8.65. No significant difference was found t(512) = -.85, p = 397. This indicates that the theta means in the two groups did not distinguish effectively between low scorers and high scorers on the assessment. This supports findings of the Boxum et al. and Arns et al. research with a very large archival sample.
As a result of this finding we decided to check the TBR at the Cz location since it had been reported by Lubar that the TBR was more robust at discriminating ADHD at this location. Using the same subject sample we found there was no significant difference between the group means of 1.71 and 1.75, t(534) = -.934, p = 351.
In contrast to this analysis we inspected the same sample set on a measure of impulsivity using the CEC dimension of impulsivity. Scores of the lowest third on impulsivity were compared with scores on impulsivity from the highest third as was done with attention. The mean of the low impulsivity group was 8.37 and the mean of the hi impulsivity group was 12.19. A analysis of subject samples of n = 1839 resulted in a highly significant difference between the two groups t(139) = 12.5, p = 149E-24.
We ran a third comparison with delta and attention. It was found that those who scored in the top third on attention problems as measured by the CEC had significantly higher delta than those with in the lower third on the CEC. A subject sample of n = 1742 resulted in significant difference between the two groups t(519) = -2.4, P = .017. The mean of those who scored low on attention problems was 9.52 and the mean of those with scored hi on attention difficulties was 10.0.
Discussion
This analysis was was done with archival data and the usual limitations of archival analysis apply here. There are also pssible differences not accounted for based on gender. There were possible variations in EEG sample length that did exist. The data is of high quality because it was collected by well trained therapists and was auto-artifacted in real time in the collection process. We utilized raw data in the analysis and it was not biased by any variations in qEEG norms. The CEC is statistically normed and cross-validated with high quality instruments with a long clinical history such as the Beck Inventories, The Test of Variables of Attention, and The MicroCog neurocognitive battery. In addition the results of the analysis are in agreement with two decades of in depth research on ADHD which further validates the accuracy of the assessments in the NewMind database.
The results confirm that attention is not related to frontal theta amplitude and that instead impulsivity is highly correlated with frontal theta amplitude. The TBR at Cz also was not significantly different between those scoring high on reported attention problems and those scoring low on reported attention problems. Of special interest was that those subjects scoring high on attention problems had a TBR of only 1.7 which suggested that the subjects with attention problems varied quite widely in their EEG patterns as also reported in the more recent literature. We also found evidence suggesting frontal delta was correlated with attention difficulties. It is well documented that high amplitude delta is correlated sleep issues as well as TBI (Ianof & Anghinah, 2017; Thatcher, 1998) and sleep quality is related to TBI recovery (Wickwire, 2020). At the same time we did not control for alpha peak frequency which could be considered a weakness in this analysis but the quality of the CEC may play a part in accurately distinguishing between subjects with attention difficulties and subjects with impulsivity problems with regard to their underlying EEG profiles. On the other hand impulsivity can impede attention to some extent but that has yet to be determined by future studies. None-the-less, the distinction remains between the two groups and is an important factor in deciding on EEG protocols as those with high amplitude theta that respond best to theta downtraining and those with slowed alpha respond best to SMR training. That additional information is a significant benefit to the clinician.
References
Arns, M., Gunkelman, J., Breteler, M., & Spronk, D. (2008). EEG Phenotypes predict treatment outcome to stimulants in children with ADHD. Journal of Integrative Neuroscience, 07(03), 421–438. https:// doi. org/ 10. 1142/ s0219 63520 80018 97
Arns, M., Conners, C. K., & Kraemer, H. C. (2013). A Decade of EEG Theta/Beta Ratio Research in ADHD. Journal of Attention Disorders,17(5), 374–383. https:// doi. org/ 10. 1177/ 10870 54712 460087
Arns, M., Loo, S. K., Sterman, M. B., Heinrich, H., Kuntsi, J., Asherson,
P., et al. (2016). Editorial Perspective: How should child psychologists and psychiatrists interpret FDA device approval? Caveat emptor. Journal of Child Psychology and Psychiatry, 57(5), 656–658. https:// doi. org/ 10. 1111/ jcpp. 12524
Boxum, M., Voetterl, H., van Dijk, H., Gordon, E., DeBeus, R., Arnold, L. E., & Arns, M. (2024). Challenging the Diagnostic Value of Theta/Beta Ratio: Insights From an EEG Subtyping Meta-Analytical Approach in ADHD. Applied psychophysiology and biofeedback, 10.1007/s10484-024-09649-y. Advance online publication. https://doi.org/10.1007/s10484-024-09649-y
Clarke, A. R., Barry, R. J., & Johnstone, S. (2020). Resting state EEG power research in Attention-Deficit/Hyperactivity Disorder: A review update. Clinical Neurophysiology, 131(7),1463–1479.
Clarke, A. R., Barry, R. J., McCarthy, R., & Selikowitz, M. (1998). EEG analysis in Attention-Deficit/Hyperactivity Disorder: A comparative study of two subtypes. Psychiatry Research, 81(1), 19–29. https:// doi. org/ 10. 1016/ s0165- 1781(98) 00072-9
Duclos, C., Dumont, M., Arbour, C., Paquet, J., Blais, H., Menon, D. K., De Beaumont, L., Bernard, F., & Gosselin, N. (2017). Parallel recovery of consciousness and sleep in acute traumatic brain injury. Neurology, 88(3), 268–275. https://doi.org/10.1212/WNL.0000000000003508
Kerson, C., deBeus, R., Lightstone, H., Arnold, L. E., Barterian, J., Pan, X., & Monastra, V. J. (2019). EEG Theta/Beta Ratio Calculations Differ Between Various EEG Neurofeedback and Assessment
Software Packages: Clinical Interpretation. Clinical EEG and Neuroscience, 51(2), 114–120. https:// doi. Org/ 10. 1177/ 1550059419 888320
Ianof, J. N., & Anghinah, R. (2017). Traumatic brain injury: An EEG point of view. Dementia & neuropsychologia, 11(1), 3–5. https://doi.org/10.1590/1980-57642016dn11-010002
Lansbergen, M. M., Arns, M., van Dongen-Boomsma, M., Spronk, D., & Buitelaar, J. K. (2011). The increase in theta/beta ratio on resting-state EEG in boys with attention-deficit/hyperactivity
disorder is mediated by slow alpha peak frequency. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 35(1), 47–52. https:// doi. org/ 10. 1016/j. pnpbp. 2010. 08. 004
Lubar, J. F. (1991). Discourse on the development of EEG diagnostics and biofeedback for attention-deficit/hyperactivity disorders. Biofeedbackand Self-Regulation, 16(3), 201–225.
Mohagheghi, A., Amiri, S., Moghaddasi Bonab, N., Chalabianloo, G., Noorazar, S. G., Tabatabaei, S. M., & Farhang, S. (2017). A Randomized Trial of Comparing the Efficacy of Two Neurofeedback Protocols for Treatment of Clinical and Cognitive Symptoms of ADHD: Theta Suppression/Beta Enhancement and Theta Suppression/Alpha Enhancement. BioMed research international, 2017, 3513281. https://doi.org/10.1155/2017/3513281
Monastra, V. J., Lubar, J. F., Linden, M., VanDeusen, P., Green, G.,Wing, W., et al. (1999). Assessing attention deficit hyperactivity disorder via quantitative electroencephalography: An initial validation study. Neuropsychology, 13(3), 424–433. https:// doi.org/ 10. 1037/ 0894- 4105. 13.3. 424
Paluch, K., Jurewicz, K., Rogala, J., Krauz, R., Szczypińska, M., Mikicin, M., Wróbel, A., & Kublik, E. (2017). Beware: Recruitment of Muscle Activity by the EEG-Neurofeedback Trainings of High Frequencies. Frontiers in human neuroscience, 11, 119. https://doi.org/10.3389/fnhum.2017.00119
Snyder, S. M., Quintana, H., Sexson, S. B., Knott, P., Haque, A. F. M., & Reynolds, D. A. (2008). Blinded, multi-center validation of EEG and rating scales in identifying ADHD within a clinical
sample. Psychiatry Research, 159(3), 346–358. https:// doi. org/ 10.1016/j. psych res. 2007. 05. 006
Snyder, S. M., Rugino, T. A., Hornig, M., & Stein, M. A. (2015). Integration of an EEG biomarker with a clinician’s ADHD evaluation. Brain and Behavior, 5(4), e00330. https:// doi. org/ 10. 1002/
brb3. 330
Soutar, R. (2014). The cognitive emotional checklist: validity & reliability. Woodstock, GA, NewMind.
Thatcher, R., Biver, C., McAlaster, R., Camacho, M., Salazarm, A. (1998). Biophysical linkage between MRI and EEG amplitude in closed head injury. Neuroimage 1998, 7(4):352-367.
van Dijk, H., deBeus, R., Kerson, C., Roley-Roberts, M. E., Monastra, V. J., Arnold, L. E., et al. (2020). Different Spectral Analysis Methods for the Theta/Beta Ratio Calculate Different Ratios But
Do Not Distinguish ADHD from Controls. Applied Psychophysiology and Biofeedback, 45(3), 165–173. https:// doi. org/ 10. 1007/s10484- 020- 09471-2
Wickwire E. M. (2020). Why sleep matters after traumatic brain injury. Journal of clinical sleep medicine : JCSM : official publication of the American Academy of Sleep Medicine, 16(S1), 5–6. https://doi.org/10.5664/jcsm.8872