Migraine Duration Linked to Left-Right Brain Diffusion Differences

In a cross-sectional case-control study published in Frontiers in Pain Research, researchers identified a specific association between the duration of reported migraine history and left-right differences in a diffusion MRI metric. The study, conducted at the First Affiliated Hospital of Chongqing Medical University, found that longer reported migraine duration was associated with a less leftward diffusion profile in periventricular white matter. This finding emerged from a detailed analysis of 89 participants, including 64 individuals with migraine and 25 healthy controls, who underwent diffusion tensor imaging. The research team utilized a specialized index known as DTI-ALPS, which estimates directionally constrained water diffusion near the lateral ventricles, to examine whether the side of the brain showed distinct patterns related to how long participants had experienced migraines.

The DTI-ALPS index serves as an indirect magnetic resonance imaging marker for fluid movement along cerebral perivascular pathways, a process that contributes to brain homeostasis. It is important to note that this index is not a direct measurement of glymphatic flow, as the resulting values are influenced by perivascular diffusivity, white-matter architecture, crossing fibres, and acquisition conditions. Previous studies on migraine have produced inconsistent patterns regarding this index, with some reporting higher right-sided values in chronic migraine and others observing reduced levels accompanied by impaired meningeal lymphatic measures. This heterogeneity in the literature motivated the current investigation, which aimed to clarify whether reported disease duration relates differently to left and right periventricular diffusion, while also examining adjusted bilateral differences between patients and controls.

The study design was a single-centre cross-sectional analysis that recruited participants between December 2021 and June 2023. The patient cohort comprised 64 individuals diagnosed with mutually exclusive subtypes of migraine: 20 with episodic migraine, 24 with chronic migraine, and 20 with vestibular migraine. These patients were frequency-matched by age and sex to 25 healthy controls who were free of lifetime headache disorders. The participants were scanned using a 3.0-T MRI system, and the imaging protocol was designed to limit acute-state effects. For example, participants with episodic migraine or vestibular migraine were scanned only when headache and vertigo had been absent for the preceding 24 hours, ensuring that the measurements reflected a stable baseline rather than an active attack state.

Statistical analysis employed generalized estimating equations to account for the paired nature of the left and right hemispheric measurements. The primary model for the patient cohort included reported disease duration, age, baseline monthly migraine days, sex, and body mass index, with each variable allowed a side-specific association. This approach enabled the researchers to test whether the slope of the relationship between duration and the diffusion index differed between the left and right sides. The focal test was the duration-by-side interaction, which quantified the difference in duration slopes after adjusting for the other covariates. The model used an exchangeable working correlation and robust sandwich covariance to handle the clustering of hemispheric observations within each participant.

The results indicated that there was no evidence of an average bilateral patient-control difference in the diffusion index. After adjustment, the estimated patient-control contrast on the left side was 0.0230, with a 95% confidence interval ranging from -0.0463 to 0.0923, and a p-value of 0.515. Similarly, the derived bilateral-average patient-control contrast was 0.0226, which also failed to reach statistical significance. These findings suggest that, on average, individuals with migraine do not exhibit a globally higher or lower diffusion index compared to healthy controls when both hemispheres are considered together. The lack of a significant group difference aligns with some previous pilot studies but contrasts with reports that found specific abnormalities in certain migraine subtypes.

In contrast to the null group difference, the analysis revealed a significant interaction between reported migraine duration and hemisphere side. The duration-by-side coefficient was 0.0053 per year, with a 95% confidence interval of 0.0022 to 0.0084 and a p-value less than 0.001. This positive interaction indicated that the relationship between duration and the diffusion index differed by side. Specifically, the adjusted duration slope was negative on the left side (-0.0047 per year, p = 0.025), meaning that longer duration was associated with lower left-sided values. On the right side, the slope was near zero (0.0006 per year, p = 0.760) and not statistically significant. Consequently, as reported migraine duration increased, the left-right difference in the diffusion index became less pronounced, or less leftward, within the migraine cohort.

The researchers conducted several sensitivity analyses to assess the robustness of these findings. They reprocessed the imaging data using different methods for handling directional maps, including scalar resampling and full tensor reorientation. The duration-by-side estimate remained positive under both processing approaches, suggesting that the finding was not an artifact of a specific image processing pipeline. Additionally, the team evaluated alternative definitions of the regions of interest used to calculate the diffusion index. The association remained positive with an external literature-derived local region definition, but it was not reproduced when using a broader atlas-derived sampling scheme. This spatial sensitivity supports a local interpretation of the association, indicating that the effect is specific to the sampled periventricular white matter regions rather than a widespread hemispheric phenomenon.

The study also examined the relationship between age and hemisphere side, finding no significant interaction (p = 0.788). This suggests that the age-related changes in the diffusion index were similar on both sides, unlike the duration-related changes which were lateralized. The correlation between reported duration and age was moderate (Pearson r = 0.487), and the statistical model included both variables to account for their shared variance. By allowing age and duration to have side-specific associations, the model provided a more nuanced view of how these clinical factors relate to brain diffusion properties. The baseline monthly migraine days, a measure of current disease frequency, did not show a significant side-specific interaction, further highlighting the distinct role of reported disease history in the observed lateralized pattern.

Limitations of the study include its cross-sectional design, which prevents causal inferences about how migraine duration leads to changes in brain diffusion. The association describes differences between participants with different durations of disease history, not change within an individual over time. The cohort was relatively small and predominantly female, which limits the precision of the estimates and the ability to make sex-specific inferences. Additionally, detailed data on headache laterality, aura, and handedness were incomplete for some participants, which restricted the ability to perform phenotype-specific analyses. The researchers noted that handedness may itself affect perivascular diffusion measures, and the lack of comprehensive handedness data is a potential confounding factor.

The findings contribute to the growing body of literature on the neuroimaging correlates of migraine, particularly regarding the role of perivascular pathways. These changes may be related to the clinical laterality often observed in migraine, where attacks frequently show a predominant side. The study underscores the importance of retaining paired local information when analyzing diffusion MRI data in headache research, as averaging the hemispheres can obscure significant side-specific associations.

The research was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University, and all participants provided written informed consent. The study was reported in accordance with the STROBE guideline for observational studies. The authors acknowledged the use of AI tools for analysis-script development and preliminary image screening, with details provided in the acknowledgements. The data and supplementary materials are available to support the findings, allowing for further scrutiny and replication by the scientific community.

The heterogeneity in previous studies may be partly explained by differences in how bilateral measurements are summarized, and this work demonstrates the value of analyzing left and right sides separately.

Subject of Research: Association between migraine duration and left-right differences in periventricular DTI-ALPS diffusion MRI metrics

Article Title: Migraine duration and left-right differences in periventricular DTI-ALPS: a case-control diffusion MRI study

Article References: Hu, W., Che, L., Huang, Q., Xu, Y., Luo, S., Li, H., Fan, X., Dong, L., & Zhou, J. (2026). Migraine duration and left-right differences in periventricular DTI-ALPS: a case-control diffusion MRI study. Frontiers in Pain Research, 7, 1993452. https://doi.org/10.3389/fpain.2026.1993452

Image Credits: AI Generated

DOI: 10.3389/fpain.2026.1993452

Keywords: migraine, diffusion MRI, DTI-ALPS, periventricular white matter, case-control study, brain imaging, neurology, headache research, glymphatic system, clinical association, duration, left-right