Acute Effects of Thoracolumbar Fascia Myofascial Release on Cortical Activity (MFR-EEG)

The thoracolumbar fascia (TLF) is increasingly recognized not merely as a passive connective tissue structure, but as a neurophysiologically relevant tissue involved in mechanical loading, proprioception, nociception, and sensorimotor integration. Biomechanical studies have shown that the lumbodorsal/thoracolumbar fascia exhibits viscoelastic properties responsive to mechanical stress, and narrative reviews have proposed this tissue as a potential source of low back pain. Schleip's neurobiological model further suggests that myofascial techniques may act not only through mechanical tissue deformation, but through stimulation of mechanoreceptors within the fascia, potentially producing measurable changes at the level of the central nervous system rather than being limited to local peripheral effects.

While clinical outcomes of myofascial release (e.g., pain, range of motion, postural parameters) have been studied more extensively, the acute cortical electrophysiological effects of a thoracolumbar-fascia-targeted myofascial technique - distinguished from a sham/superficial touch condition - remain insufficiently investigated. Electroencephalography (EEG) offers a non-invasive method with high temporal resolution to examine changes in alpha, mu, beta, and theta band activity associated with tactile stimulation, somatosensory processing, pain modulation, and sensorimotor integration following manual intervention.

This study is designed as a two-arm, parallel-group, sham-controlled, single-blind (data analyst blinded) randomized controlled trial. Acute effects will be assessed within a single session, with data collected at three time points: baseline (pre), early post-intervention (0-10 minutes), and late post-intervention (30 minutes).

Participants will be randomized 1:1 to an active treatment group or a sham control group using block randomization (block size 4), stratified by sex and physical activity level (IPAQ-short: low/moderate/high). The randomization list will be generated by an independent statistician using the 'blockrand' package in R. Allocation concealment will be achieved through sequentially numbered, opaque, sealed envelopes (SNOSE), opened in the participant's presence after baseline measurements are completed. Participants will be partially blinded, as the sham condition will be presented as a placebo intervention; the treating therapist will not be blinded; the EEG operator will be blinded where possible; and the data analyst will be blinded throughout.

Due to the conflict between the prone positioning required for the myofascial technique and the higher EEG signal quality obtained in the supine position, a "sandwich protocol" will be used: baseline EEG will be recorded in the supine position, the participant will then be repositioned prone for the intervention, and subsequently returned to supine for early and late post-intervention EEG recordings.

The active treatment group will receive a 10-minute myofascial release technique (Pilat myofascial induction crossed-hands technique or an equivalent TLF-specific method) targeting the T10-L4 thoracolumbar fascia region, applying sustained moderate pressure (1.5-2.5 kg). The sham control group will receive light surface contact (<0.5 kg) over the same region and duration, without any sliding or pressure variation, presented to participants as a placebo intervention.

EEG will be recorded using an 8-channel wireless system (Enobio 8, Neuroelectrics), 24-bit resolution, 500 Hz sampling rate, with electrodes positioned at F3, F4, Fz, C3, C4, Cz, and Pz, referenced to the left earlobe (CMS) with the right mastoid as ground (DRL); a single-use ECG electrode will be placed on the lower left rib cage for heart rate and heartbeat-evoked potential (HEP) analysis. EEG preprocessing will include 0.5-45 Hz band-pass filtering, 50 Hz notch filtering, resampling to 250 Hz, bad channel interpolation, independent component analysis (ICA) for artifact removal, and rejection of epochs exceeding ±100 µV. Power spectral density will be calculated using Welch's method (2-second windows, 50% overlap).

The primary outcome measures are sensorimotor mu rhythm power (C3, C4 channels) and posterior alpha power (Pz channel), reflecting cortical areas corresponding to TLF dermatomes. Secondary outcome measures include heart rate variability indices (RMSSD, LF/HF ratio), frontal alpha asymmetry, heartbeat-evoked potential (HEP) amplitude (Cz channel), and subjective ratings of local touch and back pain (VAS). Exploratory analyses will apply classical machine learning (SVM, Random Forest) and deep learning (EEGNet) models to classify pre- versus post-intervention EEG data and evaluate whether the active myofascial technique produces an EEG pattern distinguishable from the sham condition, beyond conventional group-mean comparisons.

Sample size was determined via power analysis assuming a conservative small-to-moderate effect size (f = 0.225) for the between-group comparison, adjusted for an assumed baseline-to-follow-up correlation of r = 0.60 (effective f = 0.281 for the ANCOVA model). This yielded a required sample of 51 participants per group (102 total); accounting for an estimated 20% dropout rate, the target sample size was set at 62 participants per group (124 total).

The primary statistical analysis will use an ANCOVA model (post-intervention value as the dependent variable; group, baseline value, sex, and physical activity level as fixed effects), following the intention-to-treat principle, with missing data handled via multiple imputation. A mixed-effects model (group × time interaction, participant as random effect) will be used as a secondary analysis, and a per-protocol analysis will be reported as a sensitivity analysis. Bonferroni correction will be applied to primary comparisons, and false discovery rate (FDR) correction to secondary and exploratory analyses; cluster-based permutation testing will be used for multichannel comparisons.