The science
Migraine, fibromyalgia, and pain amplification
In migraine, biofeedback has one of the clearest evidence bases in this field: a meta-analysis of 55 studies found a medium overall effect (d = 0.58), with the strongest gains in attack frequency (d = 0.70), and benefits remained stable over an average 17-month follow-up (Nestoriuc and Martin, 2007). In fibromyalgia, an 8-week randomized neurofeedback trial significantly improved pain severity, pain interference, overall fibromyalgia impact, sleep latency, and sustained attention versus control (Wu et al., 2021).
Low back pain and osteoarthritis
ISF neurofeedback is especially interesting in persistent musculoskeletal pain. In a double-blind randomized trial in chronic low back pain, the pgACC-targeted ISF protocol produced clinically meaningful improvement in 53% for pain severity, 80% for pain interference, and 73% for disability at 1-month follow-up (Adhia et al., 2023). In a sham-controlled knee osteoarthritis study, retention reached 91%, no serious adverse events were reported, and participants described moderate-to-high perceived effectiveness; a later EEG analysis also showed frequency-specific brain changes that correlated with pain measures (Mathew et al., 2022; Adhia et al., 2024).
Phantom pain and chronic neuropathic pain
Some chronic pain conditions are driven less by ongoing tissue damage and more by central sensitization and persistent cortical reorganization: phantom pain after surgery or amputation, central neuropathic pain, and other centralized pain syndromes. A meta-analysis of 21 EEG neurofeedback chronic pain studies reported a pooled effect on pain intensity of Cohen's d = -0.76, with 10 of those studies showing clinically meaningful pain reductions of more than 30% (Patel et al., 2020). For phantom pain specifically, sensorimotor neurofeedback was first shown to control pain by reshaping the cortical representation of the phantom region. Training that dissociates this representation reduces pain, while training that reinforces it intensifies it (Yanagisawa et al., 2016); a later randomized crossover trial in patients with chronic phantom limb pain confirmed the same mechanism (Yanagisawa et al., 2022).
Tinnitus
For tinnitus, the strongest neurofeedback result so far comes from a randomized clinical trial in which real-time fMRI neurofeedback reduced Tinnitus Handicap Inventory scores by 28.2 points at 6 months, compared with 12.1 points for CBT; at 12 months the difference was 30.0 versus 4.0 points (Gninenko et al., 2024). PBM is also starting to show a signal: in a 2025 randomized placebo-controlled trial, eight active laser treatments improved the emotional THI subscale and tinnitus loudness / perception scores, while the sham group showed no comparable placebo effect (Choi et al., 2025).
In practice, qEEG can help clarify whether the pattern is being driven more by sensory overamplification, stress-linked reactivity, poor downshifting, or disrupted recovery. A broader systems medicine view is often especially useful here, because pain and tinnitus are frequently shaped by factors beyond the symptom itself, including inflammation, musculoskeletal strain, stress physiology, and overall recovery capacity.
