Fractional Anisotropy is a More Sensitive Diagnostic Biomarker Than Mean Kurtosis for Patients with Parkinson Disease with Cognitive Dysfunction: A Diffusional Kurtosis Map Tract-Based Spatial Statistics Study

Article Figures & Data

Download figure
Open in new tab
Download powerpoint
FIG 1.
Post hoc analyses results between patients with PD-MCI and PDD. Green represents the mean FA skeleton of all subjects. Yellow and red represent regions with significant statistical values (P < . 05, TFCE-corrected).
Download figure
Open in new tab
Download powerpoint
FIG 2.
A, Voxel clusters of decreased FA, MK, and MD of the PDD group more than PD-MCI group. The colorful regions (blue, green, yellow, red) represent regions with significant statistical values (P < . 05, TFCE-corrected). B, Positive correlation between the FA of cluster 1 and the MoCA. FA has 4 independent clusters, MK has 4 independent clusters, and MD has 1 independent cluster. The bar plots represent the corresponding mean diffusion metrics for each group.
Download figure
Open in new tab
Download powerpoint
FIG 3.
The ROC curves of significantly different clusters of DKI metrics and the combined ROC model. The combined model was fitted with cluster 1, 2, and 3 of the FA map and cluster 2 and 3 of the MK map.

Table 1:

Clinical data of PD-MCI and PDD groups

	PD-MCI (n = 19)	PDD (n = 18)	t/χ²/z	P
Age of PD diagnosed (mean) (yr)	63.34 (SD, 5.53)	65.73 (SD, 5.98)	−1.262	.215^{^a}
Age of PD-MCI or PDD diagnosed (mean) (yr)	64.33 (SD, 7.96)	71.10 (SD, 6.26)	−3.060	.004^{^a}**
Sex, M/F	11/8	6/12	2.245	.134^b
Disease duration (yr)	2 (range, 2.00–3.50)	3.943 (SD, 2.59)	−2.004	.052^c
Education (mean) (yr)	11.42 (SD, 3.61)	10.22 (SD, 3.20)	1.065	.294^{^a}
H-Y stage (No.) (%)	–	–	−0.836	.403^c
1	6 (31.58)	5 (27.78)	–	–
1.5	NA	1 (5.56)	–	–
2	8 (42.11)	4 (22.22)	–	–
2.5	2 (10.53)	2 (11.11)	–	–
3	3 (15.79)	5 (27.78)	–	–
5	NA	1 (5.56)	–	–
MMSE (mean)	26.36 (SD, 1.70)	19.61 (SD, 3.98)	6.765	<.001^{^a}**
MoCA (mean)	19.89 (SD, 2.57)	14.61 (SD, 3.51)	5.230	<.001^{^a}**

Note:—** indicates significant difference; NA, not applicable; en dash, no raw data or analyzed data.
↵a Independent t test.
b χ² tests.
c Mann-Whitney test.

Table 3:

Comparison between significantly different clusters and the combined ROC curve model

Clusters	AUC	Standard Error	95% CI		Compared with the Combined Model
Clusters	AUC	Standard Error			χ²	P
FA cluster 1	0.81	0.07	0.66	0.95	3.32	.068
FA cluster 2	0.85	0.06	0.73	0.98	1.80	.180
FA cluster 3	0.76	0.08	0.60	0.92	3.77	.052
FA cluster 4	0.53	0.03	0.47	0.58	47.79	<.001
MK cluster 1	0.63	0.09	0.45	0.81	8.45	.003
MK cluster 2	0.69	0.09	0.52	0.87	6.99	.008
MK cluster 3	0.70	0.09	0.52	0.87	7.12	.007
MK cluster 4	0.68	0.09	0.50	0.86	9.01	.002
MD cluster 1	0.65	0.09	0.46	0.83	10.55	.001
Combined	0.91	0.05	0.82	1.00	–	–

Table 4:

Predictors of MoCA scores in patients with PD with cognitive dysfunction: univariate and multiple linear regression analysis

Note:—Single asterisk indicates P < .05; double asterisks, P < .005; triple asterisks, P < .001.
↵a Univariate analysis based on the complete cases without missing value.
b Multiple analyses based on imputed values in predictors.