The association of diffusion tensor MRI measures of normal appearing white matter and cognition

Objective: Median and peak height of fractional anisotropy (FA) and mean diffusivity (MD) are diffusion tensor imaging (DTI) markers used to quantify white matter microstructure changes. We examine the association of DTI histogram-derived measures in global normal appearing white matter (NAWM) and cognitive decline in patients with normal cognition and cognitive impairment no dementia from a memory clinic in Singapore. Methods: A total of 252 patients (mean age: 71.1 +/- 7.6 years, 53.2% women) were included. All patients underwent clinical assessments, a brain MRI scan at baseline, and neuropsychological assessments annually for 2 years. DTI scans were processed to obtain MD and FA histogram-derived measures. The National Institute of Neurological Disorders and Stroke and the Canadian Stroke Network harmonization neuropsychological battery were used to assess cognitive function. Linear regression models with generalised estimating equation (GEE) and logistic regression models were used to examine the association between DTI histogram measures and cognitive decline. Results: When compared to baseline, MD and FA measures at Year 2 were associated with an accelerated worsening in global cognition (allp for interaction <0.001; Year 0 vs 2, MD median:-0.29 (95%CI:-0.49,-0.09) vs-0.45 (95%CI:-0.65,-0.25); MD peak height: 0.22 (95%CI: 0.07, 0.37) vs 0.37 (95%CI: 0.21, 0.53); FA median: 0.11 (95%CI:-0.05, 0.26) vs 0.22 (95%CI: 0.07, 0.37); FA peak height:-0.14 (95%CI:-0.28, 0.00) vs-0.24 (95% CI:-0.38,-0.10);). Similar findings were observed for executive function and visuomotor speed while only MD measures predicted worsening in memory domain. Interpretation: This study shows that DTI histogram measures are associated with accelerated cognitive decline suggesting the utility of DTI as a pre-clinical marker in predicting the worsening of cognition in clinical trials.