Neurofeedback/ Quantitative EEG for ADHD diagnosis

Like all psy­chi­atric dis­or­ders, ADHD is diag­nosed based on the pres­ence of par­tic­u­lar behav­ioral symp­toms that are judged to cause sig­nif­i­cant impair­ment in an indi­vid­u­al’s func­tion­ing, and not on the results of a spe­cif­ic test. In fact, recent­ly pub­lished ADHD eval­u­a­tion guide­lines from the Amer­i­can Acad­e­my of Pedi­atrics (AAP) explic­it­ly state that no par­tic­u­lar diag­nos­tic test should be rou­tine­ly used when eval­u­at­ing a child for ADHD.

While most ADHD experts would agree that no sin­gle test could or should be used in iso­la­tion to diag­nose ADHD, there are sev­er­al impor­tant rea­sons why the avail­abil­i­ty of an accu­rate objec­tive test would be useful.

First, many chil­dren do not receive a care­ful and com­pre­hen­sive assess­ment for ADHD but are instead diag­nosed with based on eval­u­a­tion pro­ce­dures that are far from optimal.

Sec­ond, although AAP guide­lines indi­cate that spe­cif­ic diag­nos­tic tests should not be rou­tine­ly used, many par­ents are con­cerned about the lack of objec­tive pro­ce­dures in their child’s eval­u­a­tion. In fact, many fam­i­lies do not pur­sue treat­ment for ADHD because the the absence of objec­tive eval­u­a­tion pro­ce­dures leads them to ques­tion the diag­no­sis. You can read a review of an inter­est­ing study on this issue at www.helpforadd.com/2006/january.htm

For these rea­sons an accu­rate and objec­tive diag­nos­tic test for ADHD could be of val­ue in many clin­i­cal sit­u­a­tions. Two impor­tant con­di­tions would have to be met for such a test to be useful.

First, it would have to be high­ly sen­si­tive to the pres­ence of ADHD, i.e., indi­vid­u­als who tru­ly have ADHD as deter­mined by a com­pre­hen­sive eval­u­a­tion should score pos­i­tive for ADHD on the test. If the test were 100% sen­si­tive, every indi­vid­ual who has ADHD based on cur­rent diag­nos­tic cri­te­ria would score pos­i­tive on the test. As the sen­si­tiv­i­ty of a test drops, the num­ber of “false neg­a­tives” — nor­mal test results in indi­vid­u­als who tru­ly have the dis­or­der increase and its util­i­ty goes down.

Sec­ond, indi­vid­u­als who don’t have ADHD should nev­er score pos­i­tive on the test, i.e., a pos­i­tive result should occur only for indi­vid­u­als with ADHD and no one else. When a diag­nos­tic test has high speci­fici­ty, indi­vid­u­als with­out the con­di­tion rarely score pos­i­tive on the test. When speci­fici­ty is low, many indi­vid­u­als with­out the con­di­tion will score pos­i­tive and may be incor­rect­ly diag­nosed as a result. This is referred to as a “false positive”.

Although many psy­cho­log­i­cal tests yield dif­fer­ent results, on aver­age, for indi­vid­u­als with and with­out ADHD, they are not sen­si­tive or spe­cif­ic enough to be par­tic­u­lar­ly use­ful when mak­ing indi­vid­ual diag­nos­tic deci­sions. For exam­ple, a wide­ly used objec­tive test in ADHD eval­u­a­tions are Con­tin­u­ous Per­for­mance Tests (CPTs). These tests pro­vide a com­put­er­ized mea­sure of a child’s abil­i­ty to sus­tain atten­tion and refrain from impul­sive respond­ing. Although aver­age per­for­mance on CPTs for chil­dren with ADHD is below that of peers, and CPT data can be help­ful when thought­ful­ly inte­grat­ed with oth­er diag­nos­tic infor­ma­tion, these tests yields too many false pos­i­tives and false neg­a­tives to be use­ful as an “objec­tive” diag­nos­tic test for ADHD.

- Is there any­thing better? -

Sev­er­al past issues of Atten­tion Research Update have reviewed Quan­ti­ta­tive EEG, i.e., QEEG, as a diag­nos­tic aide for ADHD. The use of QEEG is based on find­ings that indi­vid­u­als with ADHD have a dis­tinc­tive pat­tern of brain elec­tri­cal activ­i­ty that is often referred to as “cor­ti­cal slow­ing”; this is char­ac­ter­ized by an ele­va­tion of low fre­quen­cy theta waves and a reduc­tion of high­er fre­quen­cy beta waves in the pre­frontal cor­tex. Theta wave activ­i­ty is asso­ci­at­ed with an unfo­cused and inat­ten­tive state while beta activ­i­ty is asso­ci­at­ed with more focused atten­tion. Thus, an ele­vat­ed theta/beta ratio reflects a less alert and more unfo­cused state.

In a QEEG test­ing, EEG data is col­lect­ed from a child or adult in a non-inva­sive pro­ce­dure that requires about 30 min­utes. The EEG data is dig­i­tized and com­put­er scored so that an indi­vid­u­al’s theta/beta ratio can be com­put­ed; this ratio is then com­pared to what is typ­i­cal for indi­vid­u­als of sim­i­lar age. When this ratio is suf­fi­cient­ly ele­vat­ed ratio — the cut-off typ­i­cal­ly used is 1.5 stan­dard devi­a­tions above aver­age which cor­re­sponds to the high­est 7% of the pop­u­la­tion — the indi­vid­ual is con­sid­ered to have the EEG mark­er for ADHD.

In past stud­ies, rough­ly 90% of indi­vid­u­als diag­nosed with ADHD based on a com­pre­hen­sive eval­u­a­tion test­ed pos­i­tive for this EEG mark­er. In con­trast, about 95% of nor­mal con­trols test­ed neg­a­tive. Thus, while not a per­fect­ly reli­able indi­ca­tor, the sen­si­tiv­i­ty and speci­fici­ty of QEEG in iden­ti­fy­ing ADHD was extreme­ly strong. You can review these stud­ies at www.helpforadd.com/2001/april.htm and www.helpforadd.com/yr2000/april.htm.

The impor­tant lim­i­ta­tion of this work was that QEEG was test­ed using indi­vid­u­als known to have ADHD and nor­mal con­trols with­out any dis­or­der. Dif­fer­en­ti­at­ing between ADHD and no dis­or­der, how­ev­er, is not the sit­u­a­tion that clin­i­cians typ­i­cal­ly face. Instead, a child is referred because of atten­tion and/or behav­ior prob­lems and the clin­i­cian must deter­mine whether these prob­lems reflect ADHD, are bet­ter explained by anoth­er dis­or­der, or do not rise to the lev­el where any diag­no­sis is appro­pri­ate. Thus, for QEEG to be use­ful in ADHD eval­u­a­tions, it must also accu­rate­ly dis­tin­guish between ADHD and oth­er disorders.

One recent­ly pub­lished pre­lim­i­nary study indi­cat­ed promis­ing find­ings in this regard. Twen­ty-six chil­dren and ado­les­cents referred to an out­pa­tient psy­chi­a­try clin­ic for atten­tion and behav­ior prob­lems received a thor­ough ADHD assess­ment con­duct­ed by a team of child psy­chi­a­trists. They also received a QEEG eval­u­a­tion. Six­teen of the 26 were deter­mined to meet DSM-IV cri­te­ria for ADHD by the psy­chi­atric team while 10 were diag­nosed with oth­er con­di­tions. Of the 16 diag­nosed with ADHD, 15 showed the QEEG mark­er for ADHD; in con­trast, none of the 10 diag­nosed with oth­er con­di­tions showed the QEEG mark­er. Thus, the QEEG test per­formed extreme­ly well. A com­pre­hen­sive review of this study can be found at www.helpforadd.com/2007/november.htm

While these results were encour­ag­ing, the sam­ple was small and from a sin­gle clin­i­cal site. This rais­es impor­tant ques­tions about the gen­er­al­iz­abil­i­ty of the find­ings that need to be addressed in a study that incor­po­rates a larg­er sam­ple drawn from mul­ti­ple clin­i­cal sites. Recent­ly, a study meet­ing these cri­te­ria was pub­lished; I believe it is one of the most inter­est­ing and impor­tant stud­ies I have seen in sev­er­al years [Sny­der et al. (2008). Blind­ed, mul­ti-cen­ter val­i­da­tion of EEG and rat­ing scales in iden­ti­fy­ing ADHD with­in a clin­i­cal sam­ple. Psy­chi­a­try Research, 159, 346–358.]

- Meth­ods -

Par­tic­i­pants — Par­tic­i­pants were 159 6–18-year olds (101 males and 58 females) tak­en by par­ents to 1 of 4 pedi­atric and psy­chi­atric clin­ics because of con­cerns relat­ed to atten­tion and behav­ior prob­lems. One hun­dred and fif­teen were chil­dren (6–11 years old) and the remain­der were ado­les­cents. There was a good rep­re­sen­ta­tion of African Amer­i­cans in the sam­ple (37%).

Psy­chi­atric Exam — At each site, chil­dren received a stan­dard­ized psy­chi­atric eval­u­a­tion that includ­ed a semi-struc­tured inter­view (the KSADS-PL) and mea­sures of func­tion­al impair­ment and dis­or­der sever­i­ty. Both par­ents and children/adolescents were inter­viewed, the opti­mal pro­ce­dure for cov­er­ing both exter­nal­iz­ing and inter­nal­iz­ing dis­or­ders. Oth­er clin­i­cal pro­ce­dures includ­ed tak­ing a med­ical his­to­ry, a devel­op­men­tal his­to­ry, and pro­vid­ing a phys­i­cal exam.

Using results of these inter­views and asso­ci­at­ed mea­sures, the clin­i­cal team per­formed a com­plete dif­fer­en­tial diag­no­sis for the pres­ence of ADHD, comor­bid con­di­tions, and oth­er childhood/adolescent dis­or­ders. Diag­nos­tic deci­sions about ADHD fol­lowed a stan­dard pro­to­col to deter­mine whether strict DSM-IV diag­nos­tic cri­te­ria were met. The team’s deter­mi­na­tion about the pres­ence or absence of ADHD was con­sid­ered the “gold stan­dard” against which diag­nos­tic deci­sions based on the results from stan­dard­ized behav­ior rat­ing scales and QEEG were compared.

Rat­ing Scales — Par­ents and teacher com­plet­ed 2 behav­ior rat­ing scales that are wide­ly used in the assess­ment of ADHD — the Con­ners Rat­ing Scale and the ADHD-IV Rat­ing Scale. Chil­dren were con­sid­ered to be pos­i­tive for ADHD if their scores on these scales exceed­ed the rec­om­mend­ed cut-off for iden­ti­fy­ing ADHD. This enabled the researchers to deter­mine how well diag­nos­tic deci­sions derived fro rat­ing scales agreed with results from the com­pre­hen­sive psy­chi­atric exam.

QEEG — EEG data was col­lect­ed on each child using stan­dard col­lec­tion pro­ce­dures by trained EEG tech­ni­cians. The theta/beta ratio com­put­ed for each child and com­pared to val­ues for age matched con­trols from a large nor­ma­tive data base. Par­tic­i­pants whose theta/beta ratio was at 1.5 stan­dard devi­a­tions above the aver­age score, i.e., rough­ly the top 7%, were con­sid­ered to show the EEG mark­er for ADHD.

It is impor­tant to note that diag­nos­tic deci­sions from the psy­chi­atric eval­u­a­tion were made with­out the team hav­ing any access to rat­ing scale or QEEG data. Thus, deci­sions made about the pres­ence or absence of ADHD from the psy­chi­atric eval­u­a­tion was not influ­enced in any way by knowl­edge of these oth­er results.

- Results -

The log­ic of this study is sim­ple and straight for­ward. The authors treat­ed results of the psy­chi­atric exam as the “gold stan­dard” for deter­min­ing which of the 159 par­tic­i­pants met diag­nos­tic cri­te­ria for ADHD. Then, they exam­ined how well results based on the behav­ior rat­ing scales and the EEG exam matched this standard.

Nine­ty-sev­en of the 159 chil­dren and ado­les­cents (61%) were diag­nosed with ADHD by the psy­chi­atric eval­u­a­tion. Six­ty-four of these chil­dren were diag­nosed with at least 1 oth­er dis­or­der, 35 had at least 2 addi­tion­al dis­or­ders, and 11 had at least 3 addi­tion­al dis­or­ders. The most com­mon comor­bidi­ties were one of the dis­rup­tive behav­ior dis­or­ders (Oppo­si­tion­al Defi­ant Dis­or­der or Con­duct Dis­or­der) which occurred in 66 of the 97, fol­lowed by an anx­i­ety dis­or­der (46 of 97), learn­ing dis­or­der (33 of 97), and mood dis­or­der (23 of 97).

Of the 62 par­tic­i­pants not meet­ing cri­te­ria for ADHD, all but 8 were diag­nosed with one of these oth­er dis­or­ders while 8 had no diagnosis.

Over­all, there­fore, this was a diverse clin­i­cal sam­ple that includ­ed the full range of psy­chi­atric dif­fi­cul­ties that clin­i­cians are called on to address.

- How accu­rate were behav­ior rat­ing scales at iden­ti­fy­ing ADHD? -

The behav­ior rat­ing scales did not per­form well. The results were as follows:

ADHD-IV Par­ent — 28% false neg­a­tives, 67% false pos­i­tives, 56% over­all accuracy.

ADHD-IV Teacher — 62% false neg­a­tives, 39% false pos­i­tive, 47% over­all accuracy.

ADHD-IV P&T com­bined — 45% false neg­a­tives, 57% false pos­i­tives, 50% over­all accuracy.

Con­ners Par­ent — 22% false neg­a­tives, 86% false pos­i­tives, 55% over­all accuracy.

Con­ners Teacher — 33% false neg­a­tives, 59% false pos­i­tives, 58% over­all accuracy.

Con­ners P&T com­bined — 28% false neg­a­tives, 81% false pos­i­tives, 53% combined.

As can be seen, the rate of false neg­a­tives (chil­dren diag­nosed with ADHD based on the psy­chi­atric eval­u­a­tion that scored below the rec­om­mend­ed ADHD cut-off on the rat­ing scale) ranged from 28% for the par­ent ver­sion of the Con­ners Rat­ing Scales to 62% for the teacher ver­sion of the ADHD-IV rat­ing scale.

False pos­i­tive rates (chil­dren with­out ADHD based on the psy­chi­atric eval­u­a­tion who scored pos­i­tive for ADHD on the rat­ing scale) ranged from 39% to 86%. The extreme­ly high false pos­i­tive rate for the par­ent ver­sion of the Con­ners indi­cates that par­ents tend­ed to rate their child high on ADHD symp­toms even when ADHD was not judged to be present.

Over­all clas­si­fi­ca­tion accu­ra­cy — how often rat­ing scale results agreed with psy­chi­atric eval­u­a­tion results — was below 60% for every scale. Thus, the lev­el of agree­ment was not much bet­ter than chance.

- QEEG Accuracy -

The accu­ra­cy of QEEG as a diag­nos­tic test was much high­er — the false neg­a­tive rate was only 13% and the false pos­i­tive rate was only 6%; this result­ed in an over­all accu­ra­cy rate of 89%. These fig­ures indi­cate the following:

- 87% of chil­dren diag­nosed with ADHD by the psy­chi­atric eval­u­a­tion showed the EEG mark­er for the disorder;

- 94% of chil­dren with­out ADHD screened neg­a­tive for the EEG marker;

- If ADHD diag­no­sis was based strict­ly on the pres­ence or absence of the EEG mark­er, it would match deci­sions based on the psy­chi­atric eval­u­a­tion almost 90% of the time.

Over­all, these results are far supe­ri­or to the clas­si­fi­ca­tion accu­ra­cy using rat­ing scales.

- Extend­ing the find­ings to dif­fer­ent sub­groups and comorbidities -

Because par­tic­i­pants rep­re­sent­ed a diverse clin­i­cal sam­ple, the researchers could test whether QEEG accu­ra­cy was sim­i­lar for chil­dren vs. ado­les­cents, whites vs. blacks, and males vs. females. Across these dif­fer­ent demo­graph­ic groups, over­all accu­ra­cy rates ranged from 87% to 95%. Thus, QEEG worked well with­in all demo­graph­ic groups.

The authors also exam­ined whether diag­nos­tic accu­ra­cy of the QEEG was con­sis­tent depend­ing on whether or not oth­er dis­or­ders were present. When the psy­chi­atric eval­u­a­tion indi­cat­ed ADHD, the QEEG was equal­ly like­ly to be pos­i­tive regard­less of what oth­er psy­chi­atric con­di­tions were diag­nosed. Sim­i­lar­ly, when ADHD was not diag­nosed, the like­li­hood that the QEEG mark­er was neg­a­tive did not depend on what oth­er con­di­tions were present. Thus, the clas­si­fi­ca­tion accu­ra­cy of QEEG as a diag­nos­tic test for ADHD was not influ­enced by the pres­ence or absence of oth­er psy­chi­atric conditions.

- Sum­ma­ry and Implications -

I believe these find­ings are very impor­tant. One clear impli­ca­tion is that results from behav­ior rat­ing scales must be used cau­tious­ly in ADHD eval­u­a­tions. Mak­ing diag­nos­tic deci­sions based on count­ing symp­toms and/or deter­min­ing whether a child’s rat­ing scale results falls in a clin­i­cal­ly ele­vat­ed range for ADHD will lead to high rates of mis­di­ag­no­sis in com­par­i­son to what would emerge from a com­pre­hen­sive psy­chi­atric eval­u­a­tion. In par­tic­u­lar, data from this study sug­gests that many chil­dren who do not tru­ly have ADHD would be erro­neous­ly diag­nosed with the disorder.

I should empha­size that ADHD eval­u­a­tion guide­lines from the Amer­i­can Acad­e­my of Pedi­atrics and the Amer­i­can Acad­e­my of Child and Ado­les­cent Psy­chi­a­try clear­ly indi­cate that rat­ing scales should nev­er be used in iso­la­tion to diag­nose an indi­vid­ual with ADHD. Thus, using rat­ing scales in this way is incon­sis­tent with best prac­tice guide­lines. How­ev­er, giv­en the lim­it­ed time avail­able in many pri­ma­ry care set­tings to con­duct com­pre­hen­sive eval­u­a­tions, it would not be sur­pris­ing if rat­ing scale results are some­times giv­en greater empha­sis than is recommended.

This is where QEEG results can be so help­ful. In regards to match­ing ADHD diag­nos­tic deci­sions that result from a com­pre­hen­sive eval­u­a­tion, this test per­formed quite well — over­all accu­ra­cy rates were near­ly 90%. The false pos­i­tive rate of only 6% means that few chil­dren and ado­les­cents for whom ADHD is ruled out by a com­pre­hen­sive psy­chi­atric exam would be diag­nosed if QEEG find­ings were used to make the deci­sion. This is a strik­ing con­trast to false pos­i­tive rates that exceed­ed 80 % for some of the rat­ing scale mea­sures. Although adjust­ing how rat­ing scale data is used to make deci­sions about ADHD might improve clas­si­fi­ca­tion accu­ra­cy some­what, it is dif­fi­cult to imag­ine that over­all accu­ra­cy rates would ever approach that found for QEEG.

It is impor­tant to under­score that despite the strong results found for QEEG, this tool is not a sub­sti­tute for a com­pre­hen­sive diag­nos­tic exam and should not be used as a stand alone test for ADHD. One essen­tial rea­son for this is that diag­nos­tic eval­u­a­tions for ADHD should go beyond sim­ply decid­ing whether ADHD is present and gath­er oth­er infor­ma­tion that is crit­i­cal for devel­op­ing an opti­mal treat­ment plan. While QEEG may help with the for­mer, it does not con­tribute to the lat­ter, beyond iden­ti­fy­ing indi­vid­u­als for whom med­ica­tion treat­ment would be an appro­pri­ate option to con­sid­er. I have heard expe­ri­enced users of QEEG dis­cuss that oth­er data from the pro­ce­dure can be use­ful in more expan­sive treat­ment plan­ning, but I am not famil­iar with research that sup­ports this broad­er use.

Giv­en this lim­i­ta­tion, what val­ue would there be to rou­tine­ly incor­po­rat­ing QEEG into ADHD diag­nos­tic eval­u­a­tions. Sev­er­al things come to mind includ­ing the following:

1. In pri­ma­ry care set­tings where a com­pre­hen­sive psy­chi­atric exam can be dif­fi­cult to pro­vide, results from a 30–40 minute QEEG pro­ce­dure can iden­ti­fy with rea­son­ably high accu­ra­cy indi­vid­u­als for whom ADHD is like­ly to be an appro­pri­ate diagnosis.

2. Because false pos­i­tive rates are so low, QEEG could reduce the num­ber of indi­vid­u­als who are per­haps inap­pro­pri­ate­ly treat­ed with ADHD med­ica­tion if physi­cians referred indi­vid­u­als with neg­a­tive results for fur­ther evaluation.

3. In cas­es where par­ents are reluc­tant to pur­sue treat­ment for their child because of con­cerns that objec­tive eval­u­a­tion pro­ce­dures were lack­ing, QEEG pro­vides an objec­tive bio­log­i­cal mark­er of ADHD that can increase par­ents’ con­fi­dence in their child’s evaluation.

4. For peo­ple who con­tin­ue to doubt that ADHD is an actu­al con­di­tion with impor­tant bio­log­i­cal under­pin­nings, these find­ings high­light that the vast major­i­ty of indi­vid­u­als meet­ing DSM-IV cri­te­ria for ADHD have a dis­tinc­tive pat­tern of brain EEG activity.

Over­all, find­ings from this care­ful­ly con­duct­ed study make an impor­tant con­tri­bu­tion to doc­u­ment­ing the util­i­ty of QEEG as an objec­tive test to assist in the diag­no­sis of ADHD. If this pro­ce­dure were to become more wide­ly used, these data sug­gest that the num­ber of chil­dren and ado­les­cents who are inap­pro­pri­ate­ly diag­nosed and treat­ed for the dis­or­der would dimin­ish substantially.

Rabiner_David– 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.

For relat­ed read­ing, you may enjoy:

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

- More on Neu­ro­feed­back­’s Brain Train­ing Value

2 Comments

  1. chris harper on January 18, 2009 at 1:41

    Thank you for the excel­lent infor­ma­tion regard­ing objec­tive test­ing of ADHD. Hope­ful­ly, there will be fur­ther devel­op­ment in tests sim­i­lar to the QEEG and the psy­chi­atric com­mu­ni­ty will be able to fend off those that doubt that ADHD exists.



  2. Clare Albright, Psy.D. on October 27, 2009 at 11:12

    Impor­tant arti­cle, well writ­ten Dr. Rabin­er. Thanks. The accu­ra­cy rate for the QEEG trumps every oth­er method! I had­n’t real­ized how impor­tant the Sny­der, 2008 study is. Thank you.



About SharpBrains

SHARPBRAINS is an independent think-tank and consulting firm providing services at the frontier of applied neuroscience, health, leadership and innovation.
SHARPBRAINS es un think-tank y consultoría independiente proporcionando servicios para la neurociencia aplicada, salud, liderazgo e innovación.

Top Articles on Brain Health and Neuroplasticity

Top 10 Brain Teasers and Illusions

Newsletter

Subscribe to our e-newsletter

* indicates required

Got the book?