PSMD, a novel biomarker of small vessel disease, and its association with cognitive function — A comprehensive clinical validation study

Abstract

AbstractBackgroundA recent instrumental validation analysis positioned peak‐width of skeletonized mean diffusivity (PSMD) as a biomarker for vascular contributions to cognitive impairment and dementia (VCID) with excellent reliability, repeatability, and reproducibility. As the next step of biomarker development, the current study aimed to (1) perform a clinical validation of PSMD and cognitive function in MarkVCID and three independent replication samples, and (2) assess whether PSMD explains cognitive function above and beyond white matter hyperintensities (WMH).MethodThe clinical validation of PSMD included n = 395 participants from the multi‐site MarkVCID consortium, n = 6172 from population‐based CHARGE cohorts, n = 287 from RUSH, and n = 435 from the UC Davis ADRC cohort spanning diverse ages and racial/ethnic backgrounds. PSMD was derived from DTI using an automated algorithm and further log‐transformed to normalize its distribution. A composite measure of general cognitive function was calculated from neuropsychological tests assessing at least three different cognitive domains. Linear regression models were run to determine the association between PSMD and cognitive function, adjusting for age, sex, and education. A secondary model was adjusted for vascular risk factors: hypertension, diabetes, and smoking. Lastly, both models included WMH volume to evaluate PSMD beyond WMH.ResultHigher PSMD values were associated with lower general cognitive function in the MarkVCID (Beta (95% CI), ‐0.82 (‐1.03, ‐0.61), p<0.001) in primary models, and remained unchanged after additional adjustment for vascular risk factors in secondary model (‐0.87 (‐1.09, ‐0.65), p<0.001). These findings were replicated across the CHARGE, RUSH, and UC Davis ADRC cohorts (Table 1). We further observed that PSMD explained an additional 0.2% of the variance in cognitive function beyond WMH in the youngest cohort (48.1 ±8.9 years), whereas the variance explained rose to 2.51% for the oldest cohort (76.4 ±5.2 years).ConclusionThis comprehensive clinical validation study suggests that PSMD is related to general cognition across diverse samples, potentially explaining more variation in cognitive function than a classic cerebrovascular marker such as WMH. Together, our instrumental and clinical validation studies support using PSMD as a robust biomarker with potential for risk stratification and disease monitoring in multi‐site clinical trials of VCID. Additional longitudinal validation studies are underway.