New Study Supports Neurofeedback Treatment for ADHD

Neu­ro­feed­back — also known as EEG Biofeed­back — is an approach for treat­ing ADHD in which indi­vid­u­als are pro­vid­ed real-time feed­back on their brain­wave pat­terns and taught to alter their typ­i­cal EEG pat­tern to one that is con­sis­tent with a focused, atten­tive state. This is typ­i­cal­ly done by col­lect­ing EEG data from indi­vid­u­als as they focus on stim­uli pre­sent­ed on a com­put­er screen. Their abil­i­ty to con­trol the stim­uli, for exam­ple, keep­ing the smile on a smi­ley face, is con­tin­gent on main­tain­ing the par­tic­u­lar EEG state being trained. Accord­ing to neu­ro­feed­back pro­po­nents, learn­ing how to do this dur­ing train­ing gen­er­al­izes to real world sit­u­a­tions and this results in improved atten­tion and reduced hyperactive/impulsive behavior.

Neu­ro­feed­back treat­ment for ADHD has been con­tro­ver­sial in the field for many years and remains so today. Although a num­ber of pub­lished stud­ies have report­ed pos­i­tive results many promi­nent ADHD researchers believe that prob­lems with the design of these stud­ies pre­clude con­clud­ing that neu­ro­feed­back is an effec­tive treat­ment. These lim­i­ta­tions have includ­ed the absence of ran­dom assign­ment, the lack of appro­pri­ate con­trol groups, raters who are not ‘blind’ to chil­dren’s treat­ment sta­tus, and small sam­ples. For addi­tion­al back­ground, you can find a recent review I wrote on exist­ing research sup­port for neu­ro­feed­back treat­ment of ADHD — along with links to exten­sive reviews of sev­er­al recent­ly pub­lished stud­ies -: How Strong is the Research Sup­port for Neu­ro­feed­back in Atten­tion Deficits?

- Results from a New Study of Neurofeedback -

Recent­ly, a study of neu­ro­feed­back treat­ment for ADHD was pub­lished that address­es sev­er­al lim­i­ta­tions that have under­mined pri­or research [Gevensleben, et al., (2009). Is neu­ro­feed­back an effi­ca­cious treat­ment for ADHD? A ran­dom­ized con­trolled clin­i­cal tri­al. Jour­nal of Child Psy­chol­o­gy and Psychiatry.]

The study was con­duct­ed in Ger­many and began with 102 chil­dren aged 8 to 12. All had been care­ful­ly diag­nosed with ADHD and approx­i­mate­ly over 90% had nev­er received med­ica­tion treat­ment. About 80% were boys. Chil­dren were ran­dom­ly assigned to one of two treat­ment con­di­tions: 36 ses­sions of neu­ro­feed­back train­ing or 36 ses­sions of com­put­er­ized atten­tion train­ing. The com­put­er­ized atten­tion train­ing task was intend­ed to serve as the con­trol inter­ven­tion. Train­ing was con­duct­ed in two 50-minute blocks per ses­sions, with a short break in-between; chil­dren in both groups par­tic­i­pat­ed in two to three such train­ing ses­sions per week.

- Descrip­tion of Training -

Neu­ro­feed­back Train­ing — As not­ed above, neu­ro­feed­back entails pro­vid­ing chil­dren with real-time feed­back on their EEG state so that they become able to learn how to cre­ate and main­tain a state that is con­sis­tent with focused atten­tion. This is done by link­ing their abil­i­ty to con­trol what appears on the com­put­er screen to their abil­i­ty to pro­duce and main­tain the EEG state being trained. Tech­ni­cal details of the train­ing pro­to­cols are not sum­ma­rized here but were based on research find­ings sug­gest­ing the spe­cif­ic EEG dif­fer­ences between chil­dren with and with­out ADHD that train­ing should address. For exam­ple, one part of train­ing focused on teach­ing chil­dren to ele­vate their pro­duc­tion of high­er fre­quen­cy beta waves and supress the pro­duc­tion of low­er fre­quen­cy theta waves. This is based on pri­or find­ings that indi­vid­u­als with ADHD tend to have an ele­vat­ed ratio of theta to beta activ­i­ty rel­a­tive (see Neurofeedback/ Quan­ti­ta­tive EEG for ADHD diag­no­sis).

Com­put­er­ized Atten­tion Train­ing — This treat­ment was based on a pro­gram called ‘Skil­lies’, described as award-win­ning Ger­man learn­ing soft­ware that pro­vides sys­tem­at­ic exer­cis­es in visu­al and audi­to­ry per­cep­tion, vig­i­lance, sus­tained atten­tion, and reac­tiv­i­ty. It was con­sid­ered the ‘con­trol’ con­di­tion to which the results of neu­ro­feed­back train­ing was compared.

Per­form­ing well on the pro­gram requires chil­dren to sus­tain their atten­tion to a vari­ety of game-like tasks that become increas­ing­ly chal­leng­ing and that pro­vide chil­dren with fre­quent feed­back about their per­for­mance. Chil­dren thus receive extend­ed prac­tice in ‘pay­ing atten­tion’ for increas­ing peri­ods to tasks that become increas­ing­ly demand­ing and need to learn to sus­tain their atten­tion in order to do well. Unlike neu­ro­feed­back treatment,however, no direct feed­back on EEG state is provided.

- Exper­i­men­tal Controls -

As not­ed above, chil­dren were ran­dom­ly assigned to treat­ment con­di­tion, which is essen­tial when com­par­ing dif­fer­ent inter­ven­tions. Efforts were also made to make the inter­ven­tion expe­ri­ence as sim­i­lar as pos­si­ble, except for the crit­i­cal dif­fer­ence as to whether chil­dren received direct train­ing in man­ag­ing their EEG state. Thus, treat­ment for both groups entailed com­put­er-game like tasks that demand­ed atten­tion. In both con­di­tions, chil­dren were encour­aged to devel­op strate­gies to focus atten­tion and to prac­tice these strate­gies at home and school. They also received sim­i­lar amounts of atten­tion and praise for doing so.

By equal­iz­ing as many aspects of the train­ing expe­ri­ence as pos­si­ble, the researchers could attribute any out­come dif­fer­ences that emerged to crit­i­cal dif­fer­ences in the pro­grams them­selves, i.e., whether feed­back on EEG state was pro­vid­ed, as opposed to some extra­ne­ous fac­tor such as atten­tion from the exper­i­menters, time spent on a demand­ing com­put­er task, etc.

In addi­tion to these impor­tant con­trols, efforts were made to keep par­ents and teach­ers ‘blind’ to the type of train­ing chil­dren received. Thus, par­ents were only told that their child would receive either of two promis­ing com­put­er-based treat­ments for ADHD. They also did not accom­pa­ny their child into the treat­ment room to observe. Chil­dren’s teach­ers were also not informed about the child’s treat­ment. Although a num­ber of par­ents became aware of which treat­ment their child received, and per­haps some teach­ers did as well, it is not pos­si­ble to keep every­one tru­ly ‘blind’ in a study like this.

- Mea­sur­ing Treat­ment Outcomes -

The main out­come mea­sure used were par­ent and teacher rat­ings of chil­dren’s ADHD symp­toms. In addi­tion to rat­ings of core inat­ten­tive and hyper­ac­tive-impul­sive symp­toms, rat­ings were col­lect­ed on a vari­ety of oth­er behav­iors, e.g., oppo­si­tion­al behav­ior, con­duct prob­lems, emo­tion­al prob­lems, and social prob­lems. These rat­ing scales were obtained before and after treatment.

To rule out place­bo effects as an expla­na­tion for any treat­ment dif­fer­ences found, the researchers also asked par­ents about their atti­tudes towards treat­ment, how moti­vat­ed they thought their chil­dren were, and how sat­is­fied they were with their child’s treatment.

- Results -

Pre­lim­i­nary analy­sis revealed no group dif­fer­ences in par­ents’ atti­tudes towards, or sat­is­fac­tion with, their child’s treat­ment or in how moti­vat­ed they felt their child was. These fac­tors thus should not have influ­enced par­ents’ rat­ings of core symptoms.

Results of the par­ent and teacher behav­ior rat­ings indi­cat­ed the following:

1. Par­ents of chil­dren treat­ed with neu­ro­feed­back report­ed sig­nif­i­cant­ly greater reduc­tions in inat­ten­tive and hyper­ac­tive-impul­sive symp­toms than par­ents of con­trol chil­dren, i.e,. those who received com­put­er­ized atten­tion train­ing. The size of the group dif­fer­ence was in a range that would be con­sid­ered mod­er­ate, i.e., about .5 stan­dard deviations.

2. Teach­ers of chil­dren treat­ed with neu­ro­feed­back report­ed sig­nif­i­cant­ly greater reduc­tions in inat­ten­tive and hyper­ac­tive-impul­sive symp­toms than teach­ers of con­trol chil­dren. The size of the group dif­fer­ence was sim­i­lar in mag­ni­tude to that found for par­ents, about .5 stan­dard deviations.

3. Apart from these dif­fer­ences in core ADHD symp­toms, few group dif­fer­ences were found. How­ev­er, neu­ro­feed­back was asso­ci­at­ed with greater reduc­tions in par­ents’ rat­ings of oppo­si­tion­al and aggres­sive behav­ior. Teacher rat­ings for the two groups did not dif­fer on any of the remain­ing behav­ioral measures.

The results sum­ma­rized above reflect aver­age dif­fer­ences between the groups. The authors also exam­ined the per­cent­age of chil­dren in each group that were judged to derive sig­nif­i­cant ben­e­fit where this was defined as at least a 25% reduc­tion in core ADHD symp­toms. Fifty-one per­cent of chil­dren in the neu­ro­feed­back group met this thresh­old com­pared to only 26% of chil­dren in the atten­tion train­ing con­trol group. This dif­fer­ence was sta­tis­ti­cal­ly sig­nif­i­cant as well.

- Sum­ma­ry and Implications -

This was a well-designed study of neu­ro­feed­back treat­ment for ADHD that used ran­dom assign­ment, blind raters, and includ­ed an appro­pri­ate con­trol group. Results indi­cate that neu­ro­feed­back treat­ment yield­ed sig­nif­i­cant­ly greater reduc­tions in par­ent and teacher rat­ings of core ADHD symp­toms than the com­par­i­son treat­ment. Fur­ther­more, the mag­ni­tude of the reduc­tions appear large enough to be clin­i­cal­ly mean­ing­ful. Although the impact of neu­ro­feed­back treat­ment on oth­er aspects of chil­dren’s func­tion­ing was less pro­nounced, sig­nif­i­cant reduc­tions in par­ents’ rat­ings of oppo­si­tion­al behav­ior were also found.

Over­all, these find­ings add to the research sup­port for neu­ro­feed­back as a treat­ment for ADHD. How­ev­er, despite the many strengths of this study, there are con­cerns to note and rea­sons why some researchers will find a basis for crit­i­ciz­ing it. The main con­cerns — and my own take on them — include the following:

1. With­out get­ting too tech­ni­cal, some researchers will argue that the sta­tis­ti­cal tests used in this study were not ide­al and may have over­es­ti­mat­ed the advan­tages of neu­ro­feed­back treat­ment. My sense from exam­in­ing the data is that the pri­ma­ry find­ings would hold up even if more con­ser­v­a­tive sta­tis­ti­cal tests were employed. How­ev­er, it would be real­ly nice to see that done.

2. Neu­ro­feed­back is sup­posed to work by teach­ing chil­dren to trans­form their EEG state to one that is char­ac­ter­is­tic of chil­dren with­out ADHD. How­ev­er, there were no EEG mea­sures tak­en in this study. Thus, there is no way to know whether neu­ro­feed­back actu­al­ly result­ed in these hypoth­e­sized changes in EEG. While this is cer­tain­ly true, this has more to do with doc­u­ment­ing the mech­a­nism by which neu­ro­feed­back led to reduc­tions in ADHD symp­toms and has noth­ing to do with whether those reduc­tions actu­al­ly occurred.

I believe that some neu­ro­feed­back prac­ti­tion­ers would argue that this may have also dimin­ished the ben­e­fits pro­vid­ed by neu­ro­feed­back treat­ment. The rea­son for this is that train­ing was not matched to the spe­cif­ic EEG para­me­ters that need­ed to be altered for each indi­vid­ual and that addi­tion­al ben­e­fits would have accrued had this been done. Whether this is actu­al­ly the case, how­ev­er, would require addi­tion­al research to determine.

3. No mea­sures of chil­dren’s aca­d­e­m­ic func­tion­ing were col­lect­ed. Because improv­ing aca­d­e­m­ic per­for­mance is a crit­i­cal treat­ment tar­get for most chil­dren with ADHD, the absence of this data is an impor­tant study lim­i­ta­tion. There is no argu­ing with this and it is unfor­tu­nate that mea­sures of aca­d­e­m­ic per­for­mance in the class­room were not collected.

4. No long-term fol­low up was con­duct­ed. There is thus no basis for know­ing whether neu­ro­feed­back treat­ment result­ed in any endur­ing ben­e­fits. While this is cer­tain­ly a lim­i­ta­tion, it should be not­ed that nei­ther med­ica­tion treat­ment nor behav­ioral treat­ment have been shown to have endur­ing ben­e­fits after treat­ment ends. How­ev­er, one of the reput­ed advan­tages of neu­ro­feed­back is that it may result in endur­ing gains. Thus, adding a long-term fol­low up to this study would have been an impor­tant addition.

5. It is impor­tant to remem­ber that when improve­ment was defined as at least a 25% reduc­tion in core ADHD symp­toms, about 50% of chil­dren treat­ed with neu­ro­feed­back did not meet this thresh­old. Thus, many chil­dren did not derive sig­nif­i­cant ben­e­fit from this treat­ment even though the ben­e­fits aver­aged across all chil­dren were sta­tis­ti­cal­ly significant.

This is not sur­pris­ing as no treat­ment — includ­ing med­ica­tion — will help every­one. How­ev­er, the rate of non-respon­ders is less than what is typ­i­cal­ly found in con­trolled stud­ies of med­ica­tion treat­ment and this is impor­tant to remain aware of.

Despite these lim­i­ta­tions and con­cerns, my take on this study is that it rep­re­sents an impor­tant addi­tion to the research lit­er­a­ture on neu­ro­feed­back treat­ment for ADHD. In the con­text of oth­er pos­i­tive find­ings that have been report­ed for neu­ro­feed­back, it pro­vides addi­tion­al basis for regard­ing this as an extreme­ly promis­ing treat­ment approach for some chil­dren with ADHD.

– Dr. David Rabin­er is a child clin­i­cal psy­chol­o­gist and Direc­tor of Under­grad­u­ate Stud­ies in the Depart­ment of Psy­chol­ogy and Neu­ro­science at Duke Uni­ver­sity. He pub­lishes Atten­tion Research Update, an online newslet­ter that helps par­ents, pro­fes­sion­als, and edu­ca­tors keep up with the lat­est research on ADHD, and teach­es the online course  How to Nav­i­gate Con­ven­tion­al and Com­ple­men­tary ADHD Treat­ments for Healthy Brain Devel­op­ment.

Relat­ed arti­cles by Dr. Rabiner

- Neurofeedback/ Quan­ti­ta­tive EEG for ADHD diagnosis

- Promis­ing Cog­ni­tive Train­ing Stud­ies for ADHD

- Mind­ful­ness Med­i­ta­tion for Adults & Teens with ADHD

- How Strong is the Research Sup­port for Neu­ro­feed­back in Atten­tion Deficits?

- Self-Reg­u­la­tion and Barkley’s The­o­ry of ADHD