Attention and Neurofeedback
(Aufmerksamkeit und Neurofeedback)
by Thomas Fuchs
Attention-Deficit/Hyperactivity Disorder is the most common psychiatric disorder in childhood, with a minimum 3 % of children affected. Only 25 % of these children outgrow the symptoms connected with this disorder.
In Part 1 of this work, an overview is given to the mass of theoretical findings about this disorder including: history, primary symptoms, subtypes and associated problems, differential diagnosis, epidemiology, developmental course and adult outcome, etiologies and treatment.
Part 2 describes a study relying on studies and work of Lubar (1995) and Sterman (1996) who demonstrated that patients can alter their brainwaves and improve their self-regulation skills. Patients with attention-deficit/hyperactivity disorder can learn to suppress their cortical Theta activity and enhance their sensorimotor rhythm- (12-15 Hz) or Beta activity (15-18 Hz) in EEG. What follows is better attention and less hyperactivity and impulsivity in the behavior of the affected children.
In total, 22 children, primarily diagnosed with AD(H)D, according to DSM-IV, in a social-pediatric hospital, ranging from 8 to 12 years in age, were assigned to the following experimental condition, consisting of 30 45-minutes sessions of EEG-Biofeedback to enhance the SMR and/or Beta activity and suppressing Theta activity in EEG, spaced over 10 weeks. No other psychological treatment or medication was administered to any subjects. In pre- and post-treatment the same testbattery was carried out, consisting of a HAWIK-R (intelligence), T.O.V.A. (continous performance test), d2 (paper-pencil-test) and 3 IOWA Conners Behavior Rating Scales (parents and teacher).
This experimental group was compared to a control group consisting of 11 children well matched in age and sex, medicated with methylphenidate (Ritalin), the most common and most successfull treatment for children with this disorder. The same testbattery was carried out in pre- and post-treatment.
The results are complex but showed comparable results and success in both treatment-groups. The children in both conditions showed significant better attention and concentration abilities in the objective (d2 and T.O.V.A) and subjective (Conners Scales) measurements. The intelligence, especially the performance score, improved significantly in both groups.
This study is the first of this type in Germany and replicates findings described in the USA (Lubar et al., 1995; Linden et al. 1996; Rossiter & LaVaque, 1995) with improved methodology.
Follow-ups, further improved methodology and replications are necessary to bring further the neurofeedback training for ADHD in Europe.
The above figure shows clinical effect for five different therapeutic approaches (meta-analysis, cf. Saile, 1996). 1st = Behavioral techniques ; 2nd = Self instruction trainings ; 3rd = Relaxation biofeedback; 4th = Parent training; 5th = Working approach (?)
Decision tree, adapted from Doepfner (1995), used in the planning of multimodal therapy for ADHD school children.
For the inattentiveness measure was found a main effect, F(1,31)=29,4, p=.000 and post-hoc tests showed good improvement in both groups: the biofeedback children improved significantly (14 points on average) (t(21)=4,96, p=.000), similarly, the Ritalin group improved 12 points (t(10)=3,15, p=.010).
For the impulsivity measure a main effect, was found, F(1,31)=28,1, p=.000. The baseline results improved comparably in both groups: For the group of biofeedbacks (t(21)=4,98, p=.000) the results showed an average improvement around 22 points, for the Ritalin children (t(10)=2,93, p=.017) around 21 points.
The response time measure showed a significant main effect, F (1,31)=42,6, p=.000 and an interaction effect F(1,31)=8,1, p=.008. Post-hoc tests determined a significant improvement in both groups (biofeedback t(21)=3,64, p=.002; Ritalin t(10)=4,70, p=.001). In this case the larger improvement was found for the Ritalin children, with a mean increase of around 32 points, with only 12.6 points for the biofeedback group.
The variability measure is viewed to be the most important measure of the T.O.V.A. Very clear improvements, particularly ifor the biofeedback group, were found: biofeedback group, t(21)=6,75, p=.000, a mean increase of 37 points. Likewise a good improvement with the Ritalin group (t(10)=3,61, p=.005) with an improvement around 18 points.
Post-hoc tests demonstrated a significant improvement in “total character” operating speed for the biofeedback group (t(18)=3,23, p=.005; 17 %) and for the Ritalin group (t(10)=2,18, p=.054; 16 %).
Both groups made fewer errors in attention load: Biofeedback group of t(18)=1,94, p=.068; 14 % improvement. Ritalin group t(10)=1,28, p=.228, around 12 % improvement.
The” total character” measure showed a significant increase in “minus errors ” for both groups: biofeedback t(18)=3,62, p=.002, improved 19 %; Ritalin t(10)=2,44, p=.035, improved around 13 %.
No significant changes were found in this variable for either group (biofeedback t(18)=.37, p=.718; Ritalin t(10)=.60, p=.562).
Post-hoc tests showed significant behavioral changes (improvements) for both groups: the mothers of the biofeedback group rated their children as calmer and more attentive behavior (t(21)=4,12, p=.000), as did the mothers of the Ritalin children (t(21)=2,92, p=.015).
Similar ratings by the fathers: biofeedback children (t(21)=3,34, p=.003), Ritalin children (t(10)=3,78, p=.004).
The MANOVA resulted in main effects, for total intelligence (F(1,31)=11,2, p=.002) and performance IQ (F(1,31)=32,8, p=.000. No main effect, was found for verbal intelligence, nor were any interactions found. Post-hoc tests indicated significant improvements in total intelligence for both groups. The biofeedback group improved an average of 4 IQ points (t(21)=2,89, p=.009), the Ritalin group improved 3 IQ points (t(10)=2,71, p=.022).
Larger improvements showed up with performance IQ: the biofeedback children improved 5 IQ points in performance IQ (t(21)=4,64, p=.000), the Ritalin children, 6 IQ points (t(10)=3,65, p=.004).
Interestingly, no significant increases were found in verbal intelligence for either group.
The above results have been compiled into a single table and four figures, below
Significant pre/post differences are green