White matter hyperintensity regression: Fact or artifact?

White matter hyperintensities (WMH) are an MRI-based biomarker associated with aging, Alzheimer's disease, and vascular dementia. Although the volume of WMH typically increases over time (growth) for individuals, WMH volume in some cases can also decrease (regress). This suggests the presence of active brain injury recovery mechanisms. Whether WMH regression reflects a true biological phenomenon or results from imaging artifacts or measurement errors, however, remains controversial. Here, we review published reports, following PRISMA search guidelines, describing or referring to WMH regression, the methods used to detect and quantitate regression, and proposed underlying mechanisms. Of 174 reviewed articles, 31 (26 original research studies and five case reports) were identified as directly related to WMH regression. Technical factors such as differences in longitudinal scan parameters, motion artifacts, and the interval between baseline and follow-up scans can affect WMH volume measurements. These factors may lead to inaccurate conclusions if appropriate controls are not employed. Although the use of standardized and systematic measurement protocols is essential, there is strong evidence indicating that WMH regression is a robust and biologically important phenomenon that may be influenced by clinical interventions. Further studies are needed to investigate WMH regression in relation to cerebrovascular risk mitigation and other therapeutic strategies.