A randomized-controlled neurofeedback trial in adult attention-deficit/hyperactivity disorder

Abstract

Attention-deficit/hyperactivity disorder (ADHD) is a childhood onset disorder persisting into adulthood for a large proportion of cases. Neurofeedback (NF) has shown promising results in children with ADHD, but randomized controlled trials in adults with ADHD are scarce. We aimed to compare slow cortical potential (SCP)- and functional near-infrared spectroscopy (fNIRS) NF to a semi-active electromyography biofeedback (EMG-BF) control condition regarding changes in symptoms and the impact of learning success, as well as changes in neurophysiological parameters in an adult ADHD population. Patients were randomly assigned to SCP-NF (n = 26), fNIRS-NF (n = 21) or EMG-BF (n = 20). Outcome parameters were assessed over 30 training sessions (pre, intermediate, post) and at 6-months follow-up (FU) including 3 booster sessions. EEG was recorded during two auditory Go/NoGo paradigms assessing the P300 and contingent negative variation (CNV). fNIRS measurements were conducted during an n-back- as well as a Go/NoGo task. All three groups showed equally significant symptom improvements suggesting placebo- or non-specific effects on the primary outcome measure. Only when differentiating between learners and non-learners, fNIRS learners displayed stronger reduction of ADHD global scores compared to SCP non-learners at FU, and fNIRS learners showed specifically low impulsivity ratings. 30.8% in the SCP-NF and 61.9% of participants in the fNIRS-NF learned to regulate the respective NF target parameter. We conclude that some adults with ADHD learn to regulate SCP amplitudes and especially prefrontal hemodynamic activity during NF. We did not find any significant differences in outcome between groups when looking at the whole sample. When evaluating learners only, they demonstrate superior effects as compared to non-learners, which suggests specific effects in addition to non-specific effects of NF when learning occurs.

Figure 1

Diagram of participant flow.

Figure 2

Experimental setup for (a) EEG-, (b) EMG- and (c) fNIRS feedback.

Figure 3

Self-regulation of SCP amplitude by task (polarity; upregulation vs. downregulation) and condition (feedback vs. transfer).

Figure 4

Self-regulation of O₂Hb amplitude by task (polarity; upregulation vs. downregulation) and condition (feedback vs. transfer) averaged over left and right ROI channels. The figure shows the non-inverted O₂Hb amplitudes (opposite to the online feedback signal).

Figure 5

ADHD self-rating differentiating learners from non-learners in fNIRS- and SCP-NF compared to EMG-BF.

Figure 6

Clinical trajectories for primary outcome assessments ADHS-SB and WRI on the global scale as well as on inattention-, hyperactivity- and impulsivity subscales.

Figure 7

ERP plots at Fz and scalp topographies at time window 1000–1800 ms depicting contingent negative variation (CNV) at pre- and post-assessment in the SCP-, fNIRS- and EMG group. The shaded area indicates the time window in which CNV was measured.

Figure 8

P300 amplitudes averaged over all feedback groups at pre-, mid-, post- and follow-up assessment at Pz and scalp topographies at time window 250–450 ms.