Stroke is neurodegenerative¹and a major risk factor for dementia.² The rate and pattern of brain atrophy after stroke has not been widely investigated and has been limited to localised atrophy and cross-sectional studies. We examined correlations in the rate of longitudinal cortical atrophy in stroke patients, compared to healthy age-matched controls, to investigate whether patterns of neurodegeneration resembled healthy functional networks as has been shown in aging and dementia syndromes.³ Default mode network (DMN) degeneration is predictive of mild cognitive impairment and Alzheimer’s disease⁴ and we predicted this would be a key network to show atrophy after stroke. We hypothesised that the rate of cortical atrophy within the DMN would be more highly correlated in stroke patients than healthy controls, reflecting network-wide neurodegeneration. Data from the Cognition and Neocortical Volume after Stroke (CANVAS)⁵ study were analysed. Fifty-three patients (mean age 67, SD 11) were scanned 3 months and 1 year after ischaemic stroke on a Siemens 3T Tim Trio scanner with a 32 channel head coil. Fourteen healthy age-matched controls (mean age 68, SD 5) were scanned at the same time points. A structural MPRAGE volume with 160 sagittal slices with 1mm isotropic voxels (TR=1900ms; TE= 2.55ms, 9^o flip angle) and A 3D SPACE-FLAIR image with 160, 1mm thick sagittal slices (TR=6000ms, TE= 380ms, 120^o flip angle) were acquired as part of a longer imaging protocol. Images were automatically processed using the longitudinal stream⁶ in Freesurfer version 5.3 (http://freesurfer.net) which uses each time-point to create an unbiased within-subject reference template from which longitudinal cortical thickness changes can be estimated.^6,7 Stroke lesions were manually traced on the FLAIR image, verified by a stroke neurologist (AB) and converted to a surface based lesion mask. The lesion mask served to exclude stroke damaged regions from longitudinal cortical thickness estimations. Vertex-wise Pearson’s correlations in the rate of atrophy were calculated relative to a posterior cingulate seed region of interest in the stroke patients and healthy aged-matched controls. The posterior cingulate seed (taken from⁸; MNI coordinates -2, -54, 27) fell within the left hemisphere, therefore cortical thickness estimates were restricted to the left hemisphere. The statistical significance of vertex-wise group differences was estimated with 10000 random permutation tests. A height threshold of p<0.01 was used to correct for multiple comparisons; cluster-wise correction thresholds were determined using Monte Carlo Simulation.⁹Regions of the DMN including the posterior cingulate, medial prefrontal cortex, parahippocampal gyrus and superior frontal gyrus showed correlated rates of atrophy 1 year after ischaemic stroke (Fig 1). In comparison to the healthy control group, correlated atrophy was more widespread in stroke patients, encompassing regions of the DMN, including posterior cingulate, but extending to regions within the middle temporal gyrus and the insula (Fig 2). The default mode network (DMN) is particularly vulnerable to atrophy and its degeneration is evident in aging and predictive of mild cognitive impairment and Alzheimer’s disease⁴. It is becoming clear that the effects of stroke are not limited to the lesion site nor the acute phase of the ischaemic event. We provide evidence of correlations in the rate of atrophy within the DMN in stroke patients, suggesting a process of network-based degeneration. A direct comparison with healthy age-matched controls revealed correlated atrophy rates beyond the DMN within the middle temporal gyrus and the insula. This suggests the pattern of atrophy is complex on a background of atrophy as a result of normal aging. Further work will determine the specificity of network degeneration after stroke in comparison to healthy aging. Our method overcomes the inter-individual variability in cortical thickness estimates that bias cross-sectional studies which have been the mainstay of research into atrophy after stroke. By examining the rate of cortical atrophy we are able to determine the vulnerability of networks to degeneration after stroke. The rate and pattern of atrophy after stroke has potential as a biomarker for predicting post-stroke dementia. We examined longitudinal atrophy within the DMN after stroke and provide evidence of network-wide neurodegeneration after stroke.