Research Summary

Reduced Brain Entropy in Migraine Linked to Disease Burden in Resting-State fMRI Study

Edited by:
Anthony Calabro, MA

Key Highlights

  • Adults with migraine had reduced brain entropy in visual, dorsal attention, and default mode network regions compared with healthy controls.
  • Entropy reductions were most pronounced in chronic migraine and were associated with greater headache frequency and longer illness duration.
  • In chronic migraine, entropy increased relatively during attacks in multisensory integration and default mode network regions.

Migraine was associated with widespread reductions in brain signal complexity, particularly among patients with chronic migraine, according to a resting-state functional MRI study published in NeuroImage. The investigators reported reduced brain entropy in regions involved in visual processing, attentional control, multisensory integration, self-referential processing, and cognitive-emotional regulation of pain.

Brain entropy quantifies the complexity of neural dynamics, with lower entropy interpreted in this study as more constrained dynamics, and higher entropy as more adaptable dynamics. The researchers examined whether alterations in entropy were associated with migraine burden, migraine phase, and symptoms. They used the Largest Lyapunov Exponent (LLE) to characterize the chaotic dynamics underlying attack-related changes in brain complexity.

The study included adults aged 18 to 65 years with episodic migraine, chronic migraine, or no history of migraine, chronic pain, systemic medical conditions, or psychiatric conditions. The final sample included participants with episodic migraine, chronic migraine, and healthy controls. Chronic migraine participants reported a mean of 18.3 headache days per month and 27.4 years of illness duration, whereas episodic migraine participants reported 5.5 headache days per month and 12.9 years of illness duration.

Researchers acquired resting-state fMRI data and computed voxel-wise sample entropy maps after standard preprocessing. Group differences were assessed using ANCOVA, with age and sex as covariates. Associations with headache days per month, illness duration, time since the most recent attack, LLE values, and migraine-associated symptoms were assessed within affected brain regions. Recent migraine timing and symptom data were collected only in the chronic migraine cohort.

Study Findings

Four clusters showed significantly altered entropy after cluster-level family-wise error correction: the occipital cortex, the right supramarginal gyrus and superior parietal lobule, the precuneus and posterior cingulate cortex, and the medial prefrontal cortex. In each region, entropy was significantly reduced in migraine groups compared with healthy controls, with the most pronounced reductions in chronic migraine.

Lower entropy was associated with greater headache frequency in the right supramarginal gyrus and superior parietal lobule (r = −0.46; corrected P = .01), with a trend in the precuneus and posterior cingulate cortex (r = −0.36; corrected P = .063). Longer migraine duration was associated with lower entropy in the right supramarginal gyrus and superior parietal lobule (r = −0.45; corrected P = .008), precuneus and posterior cingulate cortex (r = −0.49; corrected P = .005), and medial prefrontal cortex (r = −0.50; corrected P = .005).

In chronic migraine, entropy varied with time since the most recent attack and was relatively higher during and shortly after attacks, particularly in the right supramarginal gyrus and superior parietal lobule (r = −0.55; P = .043), and in the precuneus and posterior cingulate cortex (r = −0.57; P = .034). LLE values in the right supramarginal gyrus and superior parietal lobule were negatively correlated with time since last attack (r = −0.66; P = .01), and LLE values in this cluster were significantly greater than zero (t = 5.67; P = 5.75 × 10⁻⁵).

Clinical Implications

According to the study authors, the findings suggest that migraine is characterized by reduced brain entropy, reflecting constrained neural processing and reduced adaptability in regions involved in visual processing, attentional control, self-referential processing, and cognitive-emotional regulation of pain. They also stated that transient increases in entropy during migraine attacks may reflect a partial restoration of complexity via weakly chaotic neural dynamics.

The authors noted several limitations, including the cross-sectional design, which limits inference about dynamic changes across the migraine cycle; collection of recent migraine questionnaire data only in the chronic migraine cohort; analysis of migraine with and without aura together; modest and imbalanced sample sizes; and incomplete psychiatric assessments.

They also stated that the phase dynamics and symptom-related analyses were secondary exploratory investigations without additional multiple-comparison correction and should be considered preliminary.

Expert Commentary

“This study provides compelling evidence that migraine is characterized by altered brain signal complexity, with widespread reductions in entropy during resting state, especially in chronic migraine cases,” the researchers concluded.

Reference
Saberi M, Kim DJ, Hu XS, DaSilva AF. Reduced brain entropy in migraine with partial restoration during attacks: a resting-state fMRI study. NeuroImage. 2026;333:121935. doi:10.1016/j.neuroimage.2026.12193