Attention Research Update

May 2003

"Helping parents, professionals and educators stay informed about new research on ADHD"

David Rabiner, Ph.D. Senior Research Scientist, Duke University

A Comparison of Neurofeedback and Stimulant
Medication in the Treatment of ADHD

Neurofeedback is an approach for treating ADHD that is based on findings that many individuals with ADHD show reduced EEG activity in the prefrontal cortex - i.e. cortical slowing. In neurofeedback treatment, a child (or adult) is taught to increase the production of brainwave patterns that reduce or eliminate this cortical slowing.  When this has been learned, and the pattern of cortical slowing has been eliminated, it is expected that improvements in attention and reductions in hyperactive/impulsive behavior will result.  (Note: You can a review of studies documenting a pattern of cortical slowing in individuals with ADHD at - see the article titled "An objective procedure for diagnosing ADHD?" and )

As noted in a prior issue of Attention Research Update, many ADHD researchers note that although neurofeedback treatment is consistent with current theories about the biological underpinnings of ADHD, there is insufficient scientific data documenting the efficacy of this approach. Promising results have been reported in several published studies, however.  For example, in a recently published study, it was reported that children who received neurofeedback treatment as part of a comprehensive treatment program that also included medication, behavioral parent training, and school consultation, had superior outcomes compared to children whose treatment included all these components except for neurofeedback.

Results from this study - you can read the entire review at - suggest that neurofeedback may be a useful adjunct to better established treatments - i.e. medication and behavioral treatment.  As noted in my review of this study, however, it was not possible to determine whether neurofeedback would be helpful on its own because it was always combined with medication and behavioral treatments.

In addition, a surprising result from this study was that children whose treatment did not include neurofeedback failed to show improvement according to parent and teacher ratings, even though they were receiving other treatments of proven efficacy - i.e. medication and behavior therapy.  This was perplexing and raised the question of whether the addition of neurofeedback would have been associated with better outcomes if these other treatments had yielded the expected results.

A final concern with this study is that random assignment was not used to determine which children received neurofeedback and which did not.  Instead, the decision about including neurofeedback treatment was made by parents.  Although the use of random assignment was not possible because parents were paying for treatment services, and there was no indication of pretreatment differences between children in the two groups, many researchers would argue that this prevents any definitive conclusion about the added benefits provided by neurofeedback treatment from being made.

Recently, another neurofeedback treatment study was published in which several of these issues appear to have been addressed (Fuchs, T. et al., (2003). Neurofeedback treatment for ADHD in children: A comparison with methylphenidate. Applied Pyschophysiology and Biofeedback, 28, 1-12).  Participants in this study were 34 8-12 year-old children (33 male) diagnosed with ADHD based on a semi-structured interview with parents and the child, and substandard performance on a computerized test of sustained attention.  Diagnoses were made independently by two experienced clinicians and no child had received any treatment for ADHD prior to their inclusion in the study.

Children received either neurofeedback treatment or stimulant medication treatment based on the preference of their parent.  As in the study summarized above, therefore, random assignment to treatment condition was not utilized, and the implications of this will be considered below.  Parents of 22 participants opted for their child to receive neurofeedback while the remaining 12 children were treated with medication according to their parents' wishes.

Neurofeedback treatment was conducted 3 times per week over a 12-week period.  Each session lasted between 30 and 60 minutes.  The objective was to train children to consistently produce EEG states that are associated with better sustained attention (i.e. increase the production of higher frequency beta waves and decrease the production of lower frequency alpha waves). During training sessions, children received visual and auditory feedback about their EEG state, and were rewarded for meeting pre-determined performance goals (i.e. maintaining the desired EEG state for a specific time period). When these goals were consistently met, the performance level required was increased to make the task more challenging.

All children in the medication condition were treated with methylphenidate, the generic form of Ritalin.  Most children received 3 10mg doses per day on school days only.  Over the 3-month period, however, dosage was adjusted based on parent feedback about treatment effectiveness.  As a result of these adjustments, daily dosages received by individual children ranged from 10 to 60 mg after 3 months.

Prior to treatment, all children were administered the TOVA, a computerized test of sustained attention.  A second lab-based assessment of attention called the Attention Endurance Test was also administered to participants.  This test basically required children to perform a boring clerical task where careful attention to detail is necessary to obtain a good score.

In addition to these laboratory measures of attention, parents and teacher rated children's behavior using the IOWA-Conners Behavior Rating Scale, a 10-item scale designed to measure inattentiveness, hyperactivity, and aggression.  Ratings were completed independently by mothers and fathers.  Because maternal and paternal ratings were highly similar, they were combined to produce a single parent rating score.

These same measures were administered a second time 3 months after treatment was initiated.  It is important to note that teachers were not informed about the type of treatment children were receiving.  Parents, of course, were not blind to treatment condition.  Comparing the 3-month results with those obtained before treatment started enabled the researchers to determine the impact of each treatment on children's functioning, and to see whether any significant differences between the two treatments were evident.


Prior to treatment, children in the medication and neurofeedback groups did not differ on either test of sustained attention (i.e. the TOVA or the Attention Endurance Test) or on parent and teacher behavior ratings.  Information on whether medication treated subjects were on medication at the 3-month assessment was not provided.

After 3 months of treatment, both groups showed significant improvements on the TOVA and the Attention Endurance Test.  The TOVA measures several aspects of children's performance, including attention and impulsivity.  Average scores for children in each group were clearly in the normalized range for these variables and did not differ from each other.  On the Attention Endurance Test, both groups demonstrated significant improvements in speed of performance, accuracy of performance, and the composite performance score.  The degree of improvement for children in the two groups appeared to be comparable.

Results for parent and teacher behavior ratings showed a similar pattern: prior to treatment, no significant differences were evident between children in the two groups and each group showed statistically significant and clinically meaningful reductions in ADHD symptoms at 3 months.  The reduction in symptoms reported by parents and teachers appeared to be comparable for each group.


Results from this study indicate that both stimulant medication treatment and neurofeedback treatment resulted in statistically significant and clinically meaningful improvement in children with ADHD.  These gains were evident on laboratory assessments of attention and on behavior ratings from parents and teachers.  The improvement reported by teachers is especially noteworthy because teachers - unlike parents - were not aware of which treatment children had received. Overall, the benefits obtained by children receiving each treatment appeared to be comparable.

This study adds to a growing body of literature supporting the efficacy of neurofeedback as a treatment for ADHD.  And, in comparison to the study reviewed above where neurofeedback was administered in conjunction with medication and behavioral therapy, results from this study suggest that neurofeedback can be an effective treatment on its own.

Despite these encouraging results, it is important to recognize that many researchers would argue that this study does not provide conclusive evidence for the efficacy of neurofeedback.  There are several reasons for this.

First, random assignment was not used to determine which treatment children received.  Instead, each treatment was received only by children whose parents preferred it.  On the one hand, this is what actually occurs when parents seek treatment for their child - i.e. parents decide what they think will be best rather than having a particular treatment assigned by "chance". It does, however, raise the possibility that neurofeedback would only be effective for that subset of children with ADHD whose parents are favorably predisposed to this approach.

Thus, this study does not demonstrate that neurofeedback is likely to be an effective option for children with ADHD in general.  Instead, a more appropriate conclusion - and one that was suggested by the authors - is that results indicate that neurofeedback can be an effective approach for children with ADHD whose parents have a positive attitude toward a nonpharmacological treatment.

(Note: In this particular study, the same limitation applies to what could be concluded about medication treatment - i.e. one could not conclude from this study alone that medication is effective for children with ADHD in general, but only for those children whose parents are favorably predisposed to this treatment.  However, unlike neurofeedback treatment, there is a large body of literature demonstrating the effectiveness of medication in studies where random assignment was used.)

It is also important to note that long-term follow up of children in this study was not possible, so the extent to which treatment gains may have persisted is unknown.  With medication treatment, gains typically do not persist beyond the time period that medication is actually being administered.  Proponents of neurofeedback have suggested, however, that treatment induced gains often last well beyond the actual training period.  Given the time and expense that neurofeedback can involve, this would be a very important benefit.  Whether this is the case can certainly not be determined from this study and remains an important question for subsequent reearch.

Another study limitation the authors note is that they did not actually measure whether neurofeedback treatment resulted in the types of EEG changes that were the focus of training.  Thus, it was not possible to determine whether the gains children made are actually consistent with the theory of neurofeedback treatment - i.e. that the elimination of "cortical slowing" is what caused ADHD symptoms to diminish.

Instead, it is possible that other elements associated with neurofeedback treatment such as therapist attention were an important factor in children's improvement.  Alternatively, perhaps just the experience of having to "practice" paying attention - as is required in neurofeedback treatment - is sufficient, and that obtaining feedback about one's EEG state during these practice sessions is not really necessary.

A definitive conclusion about what is actually responsible for benefits accruing from neurofeedback treatment - i.e. is it training-induced changes in EEG patterns or are other factors responsible? - would require a very complex study in which a number of other possible explanations are controlled for.  This would be an expensive and difficult undertaking, and I would be surprised if we were to learn of such a study anytime soon.

Given these gaps in the literature, what can be reasonably concluded at this time about neurofeedback as a treatment for ADHD?  Clearly, different conclusions can be reached from a careful reading of research in this area.  The conclusion I have come to is that in situations where parents prefer a non-medical treatment option, and where a child's problems are not so severe that immediate symptomatic relief is essential, there is a reasonable basis for expecting that neurofeedback treatment will prove helpful.  This seems to be the case even though we do not know exactly why this is.

In summary, this study is a nice addition to the ADHD treatment literature and is supportive of neurofeedback treatment in the circumstances described above. At the same time, however, it is important to note that as reflected in ADHD treatment guidelines recently published by the American Academy of Pediatrics, many health care professionals do not believe there is yet a strong enough evidence base to recommend neurofeedback as a first-line treatment option, and that evidence supporting the use of carefully conducted medication treatment and/or behavioral interventions is certainly stronger at this time.

Information presented in Attention Research Update is for informational purposes only, and is not a substitute for professional medical advice.  Although newsletter sponsors offer products and services that I believe will be of interest to subscribers, sponsorship of Attention Research Update does not constitute a specific endorsement or guarantee of any company's product or services.