Steven M. Snyder

A Meta-analysis of Quantitative Eeg Power Associated With Attention-deficit Hyperactivity Disorder

Summary: A meta-analysis was performed on quantitative EEG (QEEG) studies that evaluated attention-deficit hyperactivity disorder (ADHD) using the criteria of the DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, 4th edition). The nine eligible studies (N = 1498) observed QEEG traits of a theta power increase and a beta power decrease, summarized in the theta/beta ratio with a pooled effect size of 3.08 (95% confidence interval, 2.90, 3.26) for ADHD versus controls (normal children, adolescents, and adults). By statistical extrapolation, an effect size of 3.08 predicts a sensitivity and specificity of 94%, which is similar to previous results 86% to 90% sensitivity and 94% to 98% specificity. It is important to note that the controlled group studies were often with retrospectively set limits, and that in practice the sensitivity and specificity results would likely be more modest. The literature search also uncovered 32 pre–DSM-IV studies of ADHD and EEG power, and 29 of the 32 studies demonstrated results consistent with the meta-analysis. The meta-analytic results are also supported by the observation that the theta/beta ratio trait follows age-related changes in ADHD symptom presentation (Pearson correlation coefficient, 0.996, P = 0.004). In conclusion, this meta-analysis supports that a theta/beta ratio increase is a commonly observed trait in ADHD relative to normal controls. Because it is known that the theta/beta ratio trait may arise with other conditions, a prospective study covering differential diagnosis would be required to determine generalizability to clinical applications. Standardization of the QEEG technique is also needed, specifically with control of mental state, drowsiness, and medication.

Blinded, multi-center validation of EEG and rating scales in identifying ADHD within a clinical sample

Previous validation studies of attention deficit/hyperactivity disorder (ADHD) assessment by rating scales or EEG have provided Class-IV evidence per standards of the American Academy of Neurology. To investigate clinical applications, we collected Class-I evidence, namely from a blinded, prospective, multi-center study of a representative clinical sample categorized with a clinical standard. Participating males (101) and females (58) aged 6 to 18 had presented to one of four psychiatric and pediatric clinics because of the suspected presence of attention and behavior problems. DSM-IV diagnosis was performed by clinicians assisted with a semi-structured clinical interview. EEG (theta/beta ratio) and ratings scales (Conners Rating Scales-Revised and ADHD Rating Scales-IV) were collected separately in a blinded protocol. ADHD prevalence in the clinical sample was 61%, whereas the remainder had other childhood/adolescent disorders or no diagnosis. Comorbidities were observed in 66% of ADHD patients and included mood, anxiety, disruptive, and learning disorders at rates similar to previous findings. EEG identified ADHD with 87% sensitivity and 94% specificity. Rating scales provided sensitivity of 38-79% and specificity of 13-61%. While parent or teacher identification of ADHD by rating scales was reduced in accuracy when applied to a diverse clinical sample, theta/beta ratio changes remained consistent with the clinicians ADHD diagnosis. Because theta/beta ratio changes do not identify comorbidities or alternative diagnoses, the results do not support the use of EEG as a stand-alone diagnostic and should be limited to the interpretation that EEG may complement a clinical evaluation for ADHD.

Comparison of a standard psychiatric evaluation to rating scales and EEG in the differential diagnosis of attention-deficit/hyperactivity disorder.

The objective was to investigate the effectiveness of rating scales and electroencephalography (EEG) in detecting the presence of attention-deficit/hyperactivity disorder (ADHD) within a diverse clinical sample. A standard psychiatric evaluation was used to assess 26 children/adolescents who presented to a clinic because a parent suspected the presence of ADHD. EEG data was collected in a blinded protocol, and rating scales were collected as well. Although all subjects had presented with ADHD-like symptoms, only 62% were diagnosed with ADHD, while the remaining 38% had other disorders or no diagnosis. Rating scales readily classified inattentive, impulsive, and/or hyperactive symptoms as being due to ADHD, regardless of the actual underlying disorder, leading to a sensitivity of 81% and a specificity of 22%. Previous studies have observed that there is an EEG marker that identifies ADHD vs. controls, and this marker was present in 15 out of 16 of the ADHD subjects (sensitivity=94%) and in none of the subjects with ADHD-like symptoms due to other disorders (specificity=100%). In the detection of ADHD in a diverse clinical sample, rating scales and EEG were both sensitive markers, whereas only EEG was specific. These results may have important implications to ADHD differential diagnosis.

Review of clinical validation of adhd behavior rating scales

The purpose of this review is to assess the range of overall accuracies for Attention Deficit/Hyperactivity Disorder (ADHD) behavior rating scales evaluated in clinical validation studies. Studies were characterized according to the evidence standards of the American Academy of Neurology (AAN). Studies were excluded due to major design problems such as overfitting by discriminant analysis. The 13 included evaluations of rating scales revealed overall accuracy in the range of 59%–79% with a pooled mean of 69% (±7%, standard deviation) and a pooled sample size of 2,228 subjects from nine studies. While some of the excluded studies demonstrated higher overall accuracies (>79%), these studies were observed to have factors in experimental design and statistics that are known to unduly inflate accuracy. We recommend further research following the full AAN standards, namely well-designed, blinded, prospective studies of rating scales applied to clinically representative samples evaluated with a clinical standard.

Addition of EEG improves accuracy of a logistic model that uses neuropsychological and cardiovascular factors to identify dementia and MCI

To investigate whether addition of EEG would improve accuracy of a logistic model that uses neuropsychological assessment and cardiovascular history to identify dementia and mild cognitive impairment (MCI) as a single group, we collected data and constructed logistic models from a sample of 78 normal adults and 33 patients (aged 50-85 years). To determine accuracy, we compared logistic regression results to a geriatricians diagnosis of MCI or dementia that included Alzheimers disease, vascular dementia or mixed dementia. We found that the addition of EEG (non-linear complexity) to a logistic model that included both neuropsychological assessment (ADAS-Cog) and cardiovascular history increased overall accuracy from 80% to 92%. The logistic model identified dementia and MCI as a single group comprised of the following subgroups (with accuracies): Alzheimers disease (92%; 12/13), vascular dementia (73%; 8/11), mixed dementia (100%; 4/4), and mild cognitive impairment (80%; 4/5). Whereas the analysis is limited by small sample sizes and mixing of diverse pathologies, the findings do provide support that the subgroups may share changes in neuropsychological, cardiovascular, and electroencephalographic factors (specifically ADAS-Cog total score, cardiovascular history, and EEG complexity). Taken together, the study results provide support that EEG might complement the clinicians evaluation of dementia and MCI.