Neurology has issued an expression of concern over a 2013 article by a group of scientists in The Netherlands and the United States who found a potentially devastating error in their analysis.
The article, “Structural and functional brain connectivity in presymptomatic familial frontotemporal dementia,” came from the lab of John C. van Swieten, of Erasmus University in The Netherlands. According to the abstract of the article:
Objective: We aimed to investigate whether cognitive deficits and structural and functional connectivity changes can be detected before symptom onset in a large cohort of carriers of microtubule-associated protein tau and progranulinmutations.
Methods: In this case-control study, 75 healthy individuals (aged 20–70 years) with 50% risk for frontotemporal dementia (FTD) underwent DNA screening, neuropsychological assessment, and structural and functional MRI. We used voxel-based morphometry and tract-based spatial statistics for voxelwise analyses of gray matter volume and diffusion tensor imaging measures. Using resting-state fMRI scans, we assessed whole-brain functional connectivity to frontoinsula, anterior midcingulate cortex (aMCC), and posterior cingulate cortex.
Results: Although carriers (n = 37) and noncarriers (n = 38) had similar neuropsychological performance, worse performance on Stroop III, Ekman faces, and Happé cartoons correlated with higher age in carriers, but not controls. Reduced fractional anisotropy and increased radial diffusivity throughout frontotemporal white matter tracts were found in carriers and correlated with higher age. Reductions in functional aMCC connectivity were found in carriers compared with controls, and connectivity between frontoinsula and aMCC seeds and several brain regions significantly decreased with higher age in carriers but not controls. We found no significant differences or age correlations in posterior cingulate cortex connectivity. No differences in regional gray matter volume were found.
Conclusions: This study convincingly demonstrates that alterations in structural and functional connectivity develop before the first symptoms of FTD arise. These findings suggest that diffusion tensor imaging and resting-state fMRI may have the potential to become sensitive biomarkers for early FTD in future clinical trials.
Here’s the notice:
With regard to the research article “Structural and functional brain connectivity in presymptomatic familial frontotemporal dementia” by Dopper et al. (Neurology® 2013;80:814-823; published ahead of print February 6, 2013), we are publishing this expression of concern to alert readers that the authors have discovered an error in the analyses that may influence the results and are planning to resolve the issue with additional studies. We are waiting for the authors to complete the reanalysis and will update readers when the results are received.
We’ve asked the authors for comment, and will update with anything we learn.