Migraine and the white matter microstructure controversy: Time for a methodological reset?

Advances in neuroimaging have significantly enhanced our ability to study brain structure and function, offering important insights into mechanisms underlying neurological disorders. A growing number of evidence from magnetic resonance imaging (MRI) based studies suggest that migraine is associated with alterations across multiple brain networks and brain regions involved with pain processing (1). Neuroimaging studies have consistently implicated the trigeminovascular system and cortical and subcortical brain regions that are relevant for the cognitive, affective and multisensory components of the pain experience, suggesting that structural and functional alterations in these areas may underlie the broad spectrum of migraine symptoms (2).

In this issue of Cephalalgia, Gollion et al. (3) present a comprehensive investigation into the microstructural integrity of brain white matter in a large cross-sectional cohort of adults with migraine using diffusion tensor imaging (DTI). The study included 293 individuals with migraine and 154 healthy controls and applied tract-based spatial statistics (TBSS) to assess key DTI metrics, including fractional anisotropy (FA), mean (MD), axial (AD) and radial (RD) diffusivity. DTI evaluates the directional movement of water molecules in tissues, providing a non-invasive proxy for assessing white matter integrity. FA measures fiber organization, AD reflects axonal integrity, RD indicates myelination and MD captures overall diffusion magnitude (4).

Previous DTI studies have explored white matter alterations in migraine, often yielding heterogeneous findings. Many reported reduced FA, increased RD and MD, and variable changes in AD in major white matter tracts such as the corpus callosum (CC), internal capsule (IC), thalamic radiations and frontal pathways. These findings have been interpreted as possible indicators of maladaptive plasticity or subtle structural damage in pain-processing circuits (5,6). Distinct DTI profiles have also been associated with migraine subtypes. Compared to healthy controls, individuals with chronic migraine has been linked to increased RD and MD in the CC, corona radiata, IC and superior longitudinal fasciculus (7). Compared to those with migraine without aura, individuals with migraine with aura showed increased FA and decreased RD in parieto-occipital and cingulate regions. Other forms, such as menstrual and pediatric migraine, demonstrated alterations in visual, sensorimotor and thalamocortical pathways (8,9). Collectively, these findings suggest that microstructural abnormalities may differ by migraine clinical phenotype.

The findings by Gollion et al. (3) provide an important contribution to the field. In this large cohort study, no significant differences in white matter microstructure were reported between individuals with migraine and healthy controls, based on TBSS analysis of DTI metrics, including FA, MD, AD and RD. These findings were consistent across all migraine subtypes (migraine with and without aura, chronic and episodic), regardless of ictal state or pain lateralization. Notably, these results echo those of a prior, smaller study by Neeb et al., (10) which also failed to detect significant white matter abnormalities in individuals with migraine. The divergence in results between previous DTI studies and the findings of Gollion et al. (3) may be attributable, at least in part, to methodological differences, ranging from MRI acquisition protocols and preprocessing pipelines to clinical sample composition and statistical approaches. Indeed, many earlier investigations included fewer than 50 participants and lacked adequate correction for multiple comparisons. The strengths of the study by Gollion et al. (3) lie in its large sample size, rigorous statistical approach and detailed clinical phenotyping, which together offer confidence in the robustness of the negative findings. While heterogeneity remains an inherent challenge in migraine research, including in this study, the use of comprehensive subgroup analyses enhances the interpretability of the results.

The absence of detectable white matter alterations along with a lack of association between white matter integrity and headache frequency suggest that migraine disease burden may not drive changes in white matter microstructure. However, the results would have been strengthened by assessing whether changes in diffusion characteristics associated with the number of years lived with migraine, which could provide important insights into whether migraine is associated with progressive changes in white matter structure over time.

Nonetheless, the lack of detectable changes in white matter diffusion metrics using a voxelwise TBSS approach does not conclusively rule out the involvement of white matter in the migraine pathophysiology. Alternative techniques, such as tract-specific analyses or region-of-interest approaches, have identified localized changes in other studies (11) and could yield different results when applied to large datasets such as this one. Moreover, migraine is a dynamic disorder and brain structure may change over time or in response to therapy. Longitudinal imaging studies will be essential to assess whether white matter microstructural alterations emerge with disease progression or resolve with treatment.

It is also important to recognize certain limitations of the study by Gollion et al. (3). The sample was clinically heterogeneous, including patients with and without medication overuse and those receiving various preventive therapies, comprising factors that could influence diffusion characteristics. Additionally, the study lacked a targeted comparison between individuals with migraine studied during the interictal phase to those examined during ictal phases, as well as to healthy controls, limiting the ability to identify state-dependent brain changes in white matter integrity.

In summary, Gollion et al. (3) offer a methodologically rigorous contribution to our understanding of white matter integrity in migraine using DTI. Their transparent reporting of cohort characteristics (including detailed information on inclusion and exclusion criteria), study design, image acquisition and post-processing protocols, and statistical analyses exemplifies best practices in reproducible research and also aligns with the principles of the NIH HEAL Initiative, which promotes methodological clarity to enhance scientific rigor and utility (12). By adhering to these standards, their work contributes to these efforts to ensure that neuroimaging research is interpretable, replicable and maximally impactful to the field of migraine.