Neurofeedback for insomnia: Current state of research

Table 1 Study parameters for surface neurofeedback

Ref.	Groups	Age/sex	Electrode placement	Protocol	Sessions	Control/sham condition
Schabus et al[45]	16 primary insomnia; 9 insomnia misperception; 26 sleep control; 12 NF control (criteria: PSG, PSQI, DSM-IV)	Insomnia group: 38.59 ± 11.18-year-old (range: 27- to 50-year-old) / 19 females and 6 males; Sleep control: 35.52 ± 10.63 (range: 24- to 47-year-old) / 19 females and 7 males; NF control: 26.67 ± 4.46 (range: 22- to 32-year-old) / 6 females and 6 males	C3	Enhance: SMR (12-15 Hz)	12 sessions NF 12 Sham (mixed); (2-4 sessions/wk); 8 5 min blocks/session	Enhance random frequencies (other than SMR rhythm) during present sessions (½ first; ½ second)
Schabus et al[41]	24 primary insomnia (criteria: PSQI, SIS-D)	34.83 ± 10.6 (range: 24- to 46-year-old) / 17 females and 7 males	C3	Enhance: SMR (12-15 Hz)	N/A	Enhance random frequencies (other than SMR rhythm) during preset sessions
Cortoos et al[46]	9 tele-NF; 8 tele-biofeedback; 12 controls	Tele-NF: 41.5-year-old / 3 females and 6 males; Tele-biofeedback: 43.8-year-old / 3 females and 5 males; Controls: 44.4-year-old/5 females and 7 males	Cz	Enhance: SMR (12-15 Hz); Inhibit: Theta (4-8 Hz) and high beta (20-30 Hz); Biofeedback: 50 Hz	18 sessions of 1 h (2-3 sessions/wk)	Control group
Arns et al[49]	51 ADHD (SMR: 27, TBR: 10, SMR + TBR: 14); 28 controls	ADHD: (range: 6- to 53-year-old) / 14 females and 37 males; Controls: (range: 21- to 64-year-old) / 15 females and 13 males	SMR: C3;Cz or C4; TBR: Fz; FCz or Cz	Enhance: SMR (12-15 Hz); TBR (20-25 Hz and 15-20 Hz)	29-31 sessions (2-3 sessions/wk); 8 5 min blocks/session	Control group
Sungwon[51]	14 participants; NF; CBT-I (criteria: ISI)	N/A	N/A	Inhibit: Beta	N/A	CBT-I
Shin[50]	4 mild insomnia (Registered at a sleep clinic)	34.8 ± 5.3-year-old/1 female and 3 males	N/A	Enhance: Theta and sigma; Inhibit: Beta	N/A	None

ADHD: Attention deficit hyperactivity disorder; CBT-I: Cognitive behavioral treatment for insomnia; DSM-IV: Diagnostic and Statistical Manual of Mental Disorders Fourth edition; NF: Neurofeedback; PSG: Polysomnography; PSQI: Pittsburgh Sleep Quality Index; SIS-D: Structured clinical interview for sleep disorders; SMR: Sensorimotor rhythm; TBR: Theta/Beta ratio.

Table 2 Summary of the results on surface neurofeedback

Ref.	Sleep outcomes measurements	Results
Schabus et al[45]	Objective: PSG; EEG; Subjective: PSQI; WHOQLA; SSS	Significant effect for NF compared to PF: (1) Higher SMR-power directly after each sessionb; and (2) Higher spindle density for the fast spindle type (C3 only)b; Significant effect for NF and PF: (1) Reduction of subjective sleep complaints after 12 sessionsb; and (2) Unspecific increase of physical quality of lifeb; Significant effect for INS compared to MP: (1) More wake timeb; and (2) Less stage 2, stage 3 and REMb; Significant effect for MP compared to INS: Higher SMR-power after each sessionb
Schabus et al[41]	Subjective: PSQI, Sleep Diary, WHOQLA, WMS-R, BSI	Significant effect only for NF: (1) Increase of SMR activity at C3a; (2) Decrease of the number of awakeningsa; and (3) Increase of stage 3a; Significant effect for NF and PF: (1) Reduction in subjective sleep complaintsb; and (2) Increased physical quality of lifeb
Cortoos et al[46]	Objective: PSG; Subjective: Sleep diary	Significant effect only for NF: (1) Decrease of SOLa and WASOa (Sleep diaries)a; and (2) Increased TSTa and SEa (Sleep diaries); Significant effect only for BF: Increase of SEa (Sleep diaries); Significant effect for NF and PF: (1) Decrease of SOLa and WASOa (PSG); and (2) Increase of REM sleep;
Arns et al[49]	Subjective: PSQI	Significant effect only for SMR: (1) Decrease of SOLd; (2) Decrease of SOL (40.1 min to 19.1 min)b; and (3) Correlation between the change in inattention and changes in PSQIb and SOLb; Significant effect for SMR + TBR: Decrease of SOL (25.8 min to 18.8 min)a
Sungwon[51]	Objective: EEG (resting state); Subjective: ISI, PSQI, DBAS	Significant effect only for NF: (1) Decrease of ISIb; (2) Changes in PSQIa; and (3) Decreased Beta power; Significant effect only for CBT-I: (1) Decrease of ISIa; (2) Changes in PSQIa; and (3) Decrease of DBASa
Shin[50]	Objective: EEG	Significant effect only for NF: Decreased spectral power for thetaa

^aP < 0.05.

^bP < 0.01, pre vs post.

^dP < 0.01 pre-treatment vs mid-treatment.

BF: Biofeedback; BSI: Brief Symptom Inventory; DBAS: Dysfunctional Beliefs and Attitude about Sleep; EEG: Electroencephalography; INS: Primary insomnia; MP: Insomnia misperception; NF: Neurofeedback; PF: Placebo neurofeedback; PSG: Polysomnography; PSQI: Pittsburgh Sleep Quality Index; REM: Rapid eye movement; TBR: Theta/Beta ratio; SE: Sleep efficiency; SOL: Sleep onset latency; SMR: Sensorimotor rhythm; SSS: Stanford Sleepiness Scale; WASO: Wakefulness after sleep onset; WHOQLA: World Health Organization Quality of Life Assessment; WMS-R: Wechsler Memory Scale – Revised.

Table 3 Study parameters for z-score neurofeedback

Ref.	Groups	Age/sex	Electrode placement	Protocol	Sessions	Sham condition
Hammer et al[53]	5 SMR; 3 IND	49.63-year-old / 5 females and 3 males	SMR: Cz and C4; IND: 4-channel (highest abnormal sites)	SMR: Increase SMR (12-15 Hz) and inhibit excessive theta (4-8 Hz) and high beta (25-30 Hz). Normalization at Cz/C4 (Amplitude, connectivity); IND: Normalization of amplitudes, coherence, asymmetry and phase lag with the highest Z-scores (> 1.96) at the 4-channels trained	Range: 9 to 15 sessions; (2 sessions/wk); 20 min/session	None
Perez-Elvira et al[55]	6 participants with insomnia and/or learning disorder	Range: 16- to 30-year-old	N/A	Normalization of the highest Z-score	20 sessions; (2 sessions/wk)	None

IND: Individualized protocols; SMR: Sensorimotor rhythm.

Table 4 Summary of the results on z-score neurofeedback

Ref.	Sleep outcomes measurements	Results
Hammer et al[53]	Objective: QEEG (Z-score); Subjective: PSQI, ISI	Significant effect only for SMR: Decrease z-score; Significant effect only for IND: Decreased proportion of abnormal z-scores at all 19 sitesb, especially for deltab and betab bands; Significant effect for SMR and IND: (1) Decreased score for ISIb, for PSQIb: All participants finished under the threshold for insomnia (ISI < 10; PSQI < 5, except 1 at 6); and (2) Increased TSTb; Results of the 6-9 mo follow up: 5 out of 6 participants remained below the ISI threshold for insomnia; 3 of the 5 improved during the months following the end of the treatment
Perez-Elvira et al[55]	Objective: QEEG; Subjective: Visual analogue scale on their symptoms	Significant effect: (1) Significant difference of 5.70% between pre- and post-QEEGs; (2) More probable to normalize then to get further away from normalizationa; and (3) Improvement of symptoms

^aP < 0.05.

^bP < 0.01 pre vs post.

IND: Individualized protocol; ISI: Insomnia Severity Index; PSQI: Pittsburgh Sleep Quality Index; QEEG: Quantitative Electroencephalogram; SE: Sleep efficiency; SMR: Sensorimotor rhythm; TST: Total sleep time.

Table 5 Study parameters for high-resolution, relational, resonance-based, electroencephalic mirroring neurofeedback

Ref.	Groups	Age/sex	Electrode placement	Protocol	Sessions	Sham condition
Tegeler et al[56]	10 HUC; 9 UC (criteria: Clinical diagnosis (clinician referral and ISI)	HUC: 41.3 ± 17.5-year-old / 8 females and 2 males; UC: 49.5 ± 8.1-year-old / 6 females and 4 males	Two-channel, changes with each protocol	Each session comprises of 4-8 protocols	8-12 sessions of up to 90 min over 3 wk	None

HUC: HIRREM plus usual care; ISI: Insomnia Severity Index; UC: Usual care (wait-list).

Table 6 Summary of the results on high-resolution, relational, resonance-based, electroencephalic mirroring neurofeedback

Ref.	Sleep outcomes measurements	Results
Tegeler et al[56]	Subjective: ISI	Significant effects only for HUC: Decrease of ISI scoreb; Significant effect only for UC: Decrease of ISI score, once the treatment is received

^bP < 0.01 pre vs post.

HUC: HIRREM plus usual care; ISI: Insomnia Severity Index; UC: Usual care (wait-list).

Table 7 Study parameters for open-loop audiovisual training neurofeedback

Ref.	Groups	Age/sex	Electrode placement	Protocol	Sessions	Sham condition
Tang et al[37]	8 chronic insomnia (criteria: ISI)	88 ± 8.7-year-old / 7 females and 1 male	N/A	Progressive reduction from 8 Hz to 1 Hz with flickering lights	30 sessions of 30 min (at bedtime) over 4 wk	None
Tang et al[58]	9 insomnia and chronic pain AVE; 7 insomnia and chronic pain Placebo (criteria: ISI)	AVE group: 67.2 ± 5.0-year-old / N/A; Placebo group: 69.6 ± 4.4-year-old / N/A	N/A	Progressive reduction from 10 Hz to 2 Hz with pulsing lights	1 session	Progressive reduction from 1 to 0.2 Hz with constant dim light changing in color

AVE: Open-loop audiovisual training; ISI: Insomnia Severity Index.

Table 8 Summary of the results for open-loop audiovisual training neurofeedback

Ref.	Sleep outcomes measurements	Results
Tang et al[37]	Subjective: ISI, PSQI, sleep diary	Significant effect: Decreased ISI score (15.6 ± 4.8 to 8.0 ± 6.4)b
Tang et al[58]	Objective: QEEG; Subjective: ISI, PSQI	Significant effect only for AVE: Increased delta frequencies’ absolute power on all 19 EEG channels; Significant difference between AVE and placebo: Delta power at Cz

^bP < 0.01 pre vs post.

ISI: Insomnia Severity Index; PSQI: Pittsburgh Sleep Quality Index; QEEG: Quantitative electroencephalogram.

Table 9 Study parameters for brain music

Ref.	Groups	Age/sex	Electrode placement	Protocol	Sessions	Sham condition
DuRousseau et al[60]	15 OPS support; 20 first responders; 6 controls	Range 24- to 58-year-old / 13 females and 28 males	F3; F4; C3; C4	Brain Music (1-4 Hz up to 30 Hz)	N/A	Enhance random frequencies during preset sessions

OPS: Operational support; N/A: Not available.

Table 10 Summary of the results on brain music neurofeedback

Ref.	Sleep outcomes measurements	Results
DuRousseau et al[60]	Subjective: 128-questions homemade questionnaire	Significant effect for all groups (except control): Improved sleep qualitya

^aP = 0.05 pre vs post.