Article Figures & Data
Figures
- FIG 1.
FLAIR images from a subject with CAA-MCI (A) and one with a non-CAA-MCI (B), demonstrating different burdens of WMH. MD maps display the skeletonized WM tracts from the same subjects with CAA-MCI (C) and non-CAA-MCI (D). E, Histograms depict the MD values of the voxels contained in the WM tract skeleton from the same subjects with CAA-MCI (solid line) and non-CAA-MCI (dashed line). F, The boxplot represents group differences in PSMD between CAA-MCI and non-CAA-MCI. The dagger indicates the results derived from ANCOVA, adjusting for age at MRI (P < .001).
Tables
CAA-MCI (n = 24) non-CAA-MCI (n = 62) P Value Demographics Age at NPT (mean) [SD] (yr) 74.73 [5.99] 73.25 [8.33] .36 Female (No.) (%) 7 (29.2) 26 (41.9) .27 Education (median) (IQR) (yr) 16 (16, 18) 16 (14, 18) .04a Vascular risk factors Hypertension (No.) (%) 13 (54.2) 40 (64.5) .38 Diabetes (No.) (%) 1 (4.2) 10 (16.1) .17 Atrial fibrillation (No.) (%) 4 (16.7) 5 (8.1) .26 Dyslipidemia (No.) (%) 13 (54.2) 44 (71.0) .14 Neuropsychological performances MMSE (median) (IQR) 25.5 (24–28) 28 (26–29) .006a MMSE (z score) (median) (IQR) –1.88 (–3.28, –0.16) 0 (–1.48–1.11) .003a Memory (z score) (median) (IQR) –1.91 (–2.38, –0.75) –0.47 (–1.60–0.56) .005a Processing speed (z score) (median) (IQR) –0.24 (–0.62–0.15) –0.11 (–0.46–0.27) .34 Language (z score) (median) (IQR) –0.44 (–0.91–0.36) –0.36 (–1.17–0.19) .95 Executive function (z score) (median) (IQR) –0.59 (–1.89–0.22) –0.21 (–0.86–0.12) .20 Imaging PSMD (× 10–4 mm2/s) (median) (IQR) 4.48 (3.81–5.09) 3.63 (3.28–4.13) <.001a Lobar CMB count (median) (IQR) 5.5 (3.0–24.50) 0 (0–0) <.001a cSS (presence) (No.) (%) 9 (37.5) 0 (0.0) <.001a High CSO-PVS score (>2) (No.) (%) 11 (45.8) 3 (4.8) <.001a nWMHV (median) (IQR) 0.42 (0.28–1.74) 0.29 (0.14–0.75) .02a nTBV (mean) [SD] 0.61 [0.04] 0.64 [0.05] .004a Note:—NPT indicates neuropsychological tests.
↵a Significant.
- Table 2:
Association between PSMD and MRI markers of small-vessel disease in subjects with CAA-MCI and non-CAA-MCIa
Model 1b Model 2c Std. β 95% CI P Std. β 95% CI P CAA-MCI (n = 24) Lobar CMB count 0.15 –0.27 0.58 .46 0.15 –0.13 0.42 .28 cSS 0.27 –0.15 0.70 .20 0.12 –0.16 0.40 .38 CSO-PVS score 0.31 –0.09 0.71 .13 0.20 –0.10 0.49 .18 nWMHV 0.75 0.49 1.02 <.001d 0.66 0.37 0.95 <.001d nTBV –0.26 –0.79 0.26 .31 –0.17 –0.54 0.20 .34 Non-CAA-MCI (n = 62) CSO-PVS –0.02 –0.24 0.21 .87 0.10 –0.07 0.27 .25 nWMHV 0.69 0.50 0.89 <.001d 0.71 0.51 0.91 <.001d nTBV –0.23 –0.51 0.06 .11 –0.14 –0.36 0.08 .20 Note:—Std. β indicates standardized beta coefficient.
↵a Linear regression models with PSMD (×10−4 mm2/s) as the dependent variable.
↵b Simple Linear regression analyses adjusted for age.
↵c Multiple regression models, including all neuroimaging markers and adjusted for age. In subjects with non-CAA-MCI, the presence of cSS and lobar CMB count was automatically excluded from the models due to the absence of variance within the group.
↵d Significant.
- Table 3:
Association between PSMD and cognitive performance in subjects with CAA-MCI and non-CAA-MCIa
Adjusted Modelb Std. β 95% CI P Value CAA-MCI (n = 24) Global cognitive status –0.40 –1.21 0.41 .31 Memory –0.69 –1.52 0.15 .10 Processing speed/attention –1.08 –1.76 –0.40 .004c Language –0.47 –1.16 0.21 .16 Executive function –0.64 –1.37 0.10 .09 non-CAA-MCI (n = 62) Global cognitive status 0.30 –0.13 0.72 .17 Memory 0.14 –0.28 0.56 .50 Processing speed/attention 0.16 –0.27 0.59 .46 Language 0.14 –0.29 0.57 .52 Executive function 0.11 –0.30 0.53 .58 ↵a Multiple regression models with cognitive performance as the dependent variable.
↵b Models include PSMD, nWMHV, nTBV, CSO-PVS score, the presence of cSS and lobar CMB count; adjusting for the time delay between the neuropsychological evaluation and the MRI. In subjects with non-CAA-MCI, the presence of cSS and the lobar CMB count were automatically excluded from the models due to an absence of variance within the group.
↵c Significant.
