Reply: A lack of consistent brain grey matter alterations in migraine

Sir, We thank Sheng et al. (2020) for their interest in our article (Burke et al., 2020). Their letter critiques the VBM (voxel-based morphometry) meta-analysis previously conducted by Jia and Yu (2017), which was used for input coordinates for our migraine network mapping analysis. We obviously did not conduct the 2017 meta-analysis, and thus will avoid defending this paper or engaging in a debate about best practices for VBM meta-analyses. In fact, we intentionally chose an existing meta-analysis rather than conduct our own to avoid this debate, as there are many different meta-analysis techniques and no clear consensus on the ‘best’ approach (Radua and Mataix-Cols, 2012). Choosing an existing meta-analysis also has the benefit of helping us avoid any potential bias in study selection. We selected the Jia and Yu (2017) meta-analysis because it was the most recently published. Using a different VBM meta-analysis approach, Sheng and colleagues suggest a lack of consistent VBM findings in migraine. Had their meta-analysis been published at the time of our search, we would likely have used their data for our network-mapping algorithm rather than Jia and Yu (2017), as their study would have been the most recent. However, we would have modified our network mapping approach slightly to account for the lack of significant meta-analytic coordinates. When there are no significant meta-analytic findings, we perform network-mapping at the individual study level (Darby et al., 2018; Weil et al., 2019). Specifically, we use the coordinates from each individual study as an input, rather than the final coordinates from the meta-analysis. For example, we previously found that coordinates of neuroimaging abnormalities in Alzheimer’s disease map to a common brain network, even though no significant findings were identified using a conventional meta-analysis (Darby et al., 2018). We used a similar approach to study cognitive impairment and visual hallucinations in Parkinson’s disease (Weil et al., 2019). When significant meta-analytic coordinates are reported, as in Jia and Yu (2017), we can test whether the reported coordinates map to a common brain network. When there are no significant coordinates, as in the study by Sheng and colleagues, we can test whether the coordinates from each individual study map to a common brain network. Future work is needed to determine whether network mapping results using coordinates from individual studies align with network mapping results using coordinates from a significant meta-analysis. Given that the meta-analytic coordinates are derived from the individual study coordinates, we suspect the two network mapping approaches will align. Data sharing is not applicable to this article as no new data were created or analysed in this study. M.J.B. has nothing to disclose. M.D.F. has intellectual property on using connectivity imaging to guide brain stimulation but receives no royalties.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.