Neuroanatomical correlates of the CDT-14: A voxel-based morphometry study exploring incremental information beyond...

BackgroundThe Clock Drawing Test (CDT) is widely used for cognitive screening but lacks unified scoring standards and neuroanatomical validation.ObjectiveWe examined the neuroanatomical correlates and incremental validity beyond conventional CDT scoring of CDT-14, a psychometrically refined 14-item system with three subfactors (Circle, Numbering, and Hand), using voxel-based morphometry (VBM).MethodsWe analyzed 985 participants from a heterogeneous memory-clinic cohort. Whole-brain VBM compared CDT-14 total and subfactor scores with three conventional scoring systems. Models adjusted for age, sex, education level, and total intracranial volume. Incremental validity was assessed after controlling for a principal-component-derived conventional composite score, with sensitivity analyses adjusting for white matter lesion volume (WMLV).ResultsCDT-14 was associated with gray matter volume (GMV) in distributed clock-drawing regions, including the right anterior hippocampus, parahippocampal gyrus, bilateral precuneus, and prefrontal regions. Primary incremental analyses identified additional associations in bilateral ventral temporal, right occipito-temporal, and dorsal anterior cingulate regions after accounting for conventional CDT performance. These associations were attenuated after WMLV adjustment, and no clusters survived FWE correction in the WMLV-adjusted incremental model. Subfactor analyses suggested partially distinct but overlapping patterns involving frontostriatal and cingulate regions for Circle, thalamic and limbic-related regions for Numbering, and hippocampal, parietal, and occipital regions for Hand.ConclusionsCDT-14 showed cross-sectional GMV associations beyond conventional scoring in primary analyses. However, the incremental associations were not robust to WMLV adjustment at the FWE-corrected threshold and should be interpreted cautiously. Longitudinal, biomarker-informed, and externally validated studies are needed to establish predictive validity, disease specificity, and clinical utility.