Genetic Architecture of Perivascular Space Morphology in the Pediatric Brain

IntroductionPerivascular spaces (PVS) support brain homeostasis through metabolite delivery and waste clearance, yet the genetic determinants of PVS morphology during childhood remain unknown. Here, we leveraged cross-sectional Adolescent Brain Cognitive Development Study data (N = 6,600; ages 9-10), including genomics and 3T structural MRI. MethodsLinear mixed-effects models examined associations between 45 single nucleotide polymorphisms (SNPs) previously linked to adult PVS structure or function and PVS count and volume fraction (VF) across six macroregions and 28 Desikan-Killiany subregions. ResultsFifteen SNPs demonstrated significant associations with PVS macroregion morphology, predominantly VF; 21 SNPs demonstrated associations with subregion morphology. Variants near SLC13A3 showed the strongest, most widespread associations with PVS VF. Additional replicated variants implicated Wnt signaling, cell adhesion, apoptosis, and glymphatic function. ConclusionThese findings suggest genetic associations with PVS morphology are detectable in childhood, while highlighting developmental and regional specificity. Longitudinal studies are now needed to determine whether childhood PVS genetic architecture predicts trajectories of glymphatic maturation and associated cognitive outcomes. Key PointsO_LIPerivascular space morphology, mapped across 28 white matter subregions, displays substantial regional and interindividual variability during childhood. C_LIO_LISix genetic associations with perivascular morphology were only detectable at subregion resolution, demonstrating that lobar averaging can obscure spatially restricted signals. C_LIO_LINearly 50% of adult-identified perivascular space risk variants replicated in 9-10 year-olds, with anatomically patterned associations reflecting patterns of neurovascular complexity and neurodevelopmental timing. C_LI