The comparisons of the clinical and MRI measures between the iNPH and HC groups are presented in Table 1. The mean MMSE, FAB, RAVLT (immediate recall, delayed recall, and recognition), IR/FAB and DR/FAB ratios in the iNPH group were significantly lower than those in the HC group. There was no significant difference in the Re/FAB ratio between the two groups. In addition, DMN connectivity in the iNPH group was significantly reduced compared with that in the HC group, and ROI FA in the iNPH group was significantly lower than that in the HC group. The results of correlation analyses with Bonferroni correction (significance was defined for p-values < 0.05/5) are shown in Table 2. In the iNPH group, DMN connectivity was significantly positively associated with the FAB score (r = 0.613, p = 0.005) and RAVLT immediate recall score (r = 0.667, p = 0.002) and was negatively associated with the iNPHGS cognitive subscore (rs = -0.576, p = 0.010). There were no significant correlations between DMN connectivity and RAVLT delayed recall and recognition scores. In addition, ROI FA was not significantly correlated with FAB score, immediate recall, delayed recall, or recognition scores of RAVLT or iNPHGS cognitive subscore in the iNPH group.
Table 2
Results of correlation analyses in the iNPH group (n = 20).
Variables
|
DMN connectivity
|
ROI FA
|
iNPHGS cognitive
|
rs = -0.576, p = 0.010*
|
rs = -0.040, p = 0.872
|
FAB
|
r = 0.613, p = 0.005**
|
r = 0.372, p = 0.117
|
RAVLT immediate recall
|
r = 0.667, p = 0.002**
|
r = -0.160, p = 0.512
|
RAVLT delayed recall
|
r = 0.510, p = 0.026
|
r = -0.054, p = 0.826
|
RAVLT recognition
|
r = 0.552, p = 0.014
|
r = -0.290, p = 0.229
|
Pearson’s partial correlation coefficient corrected for age with Bonferroni correction were used except for the iNPHGS cognitive subscore (Spearman’s partial rank correlation coefficient).
DMN: the default mode network; FAB: the frontal assessment battery; iNPH: idiopathic normal pressure hydrocephalus; iNPHGS: the idiopathic Normal Pressure Hydrocephalus Grading Scale; ROI FA: the mean fractional anisotropy value within the brain white matter region of interest.
*p < 0.01. **p < 0.005.
Figure 3 demonstrates the scatterplot between DMN connectivity and ROI FA in the patients with iNPH. We divided the patients with iNPH into the preserved DMN connectivity group (n = 12) and reduced DMN connectivity group (n = 6). The remaining two patients were excluded from the following analyses. The area under the ROC curve of the DMN/FA ratio was 0.986, and the optimal cut-off value of the DMN/FA ratio was 6.285. Based on this cut-off value, the sensitivity was 91.7%, and the specificity was 100%. Figure 2-B demonstrates the plots of DMN/FA ratios in the preserved and reduced DMN connectivity groups. The comparisons of the baseline demographic and clinical and MRI measures between the preserved and reduced DMN connectivity groups are shown in Table 3. There were no significant differences in age, sex ratio, educational attainment, disease duration, total score and subscores of iNPHGS, MMSE and FAB scores, Re/FAB ratio, and completion time and number of steps of TUG, whereas the immediate recall, delayed and recognition scores of RAVLT, and IR/FAB and DR/FAB ratios in the reduced DMN connectivity group were significantly worse than those in the preserved DMN connectivity group.
Table 3
Comparison of data at baseline between the preserved and reduced DMN connectivity groups.
Variables
|
Preserved DMN connectivity group
(n = 12)
|
Reduced DMN connectivity group
(n = 6)
|
P value
|
Age (years)
|
80.5, 65–89
|
79.5, 73–85
|
0.892
|
Sex (women/men)
|
4/8
|
2/4
|
1.000
|
Disease duration (years)
|
2.5, 0.2–5.6
|
1.9, 0.4–5.1
|
0.553
|
Educational attainment (years)
|
|
|
|
Responder/non-responder
|
10/2
|
3/3
|
0.137
|
iNPHGS
|
|
|
|
Gait
|
2.0, 1–3
|
2.0, 0–3
|
0.750
|
Cognition
|
2.0, 1–3
|
2.0, 2–3
|
0.250
|
Urination
|
1.0, 0–3
|
2.0, 0–3
|
0.553
|
Total
|
5.0, 2–8
|
6.0, 2–8
|
0.385
|
MMSE (/30)
|
25.0, 13–29
|
23.0, 11–27
|
0.291
|
FAB (/18)
|
13.0, 9–15
|
12.5, 8–16
|
0.964
|
RAVLT
|
|
|
|
Immediate recall (/75)
|
25.5, 14–41
|
18.5, 6–21
|
0.010*
|
Delayed recall (/15)
|
2.5, 0–9
|
0.5, 0–3
|
0.041*
|
Recognition (/30)
IR/FAB
DR/FAB
Re/FAB
|
26.5, 20–30
2.12, 1.27–2.79
0.22, 0-0.6
2.07, 1.6–2.8
|
21.5, 14–28
1.27, 0.75–1.64
0.04, 0-0.2
0.88, 1-2.63
|
0.041*
0.001**
0.041*
0.250
|
TUG
|
|
|
|
Time to complete (second)
|
12.2, 7.2–53.7
|
14.7, 8.4–50.2
|
0.490
|
Number of steps
DMN functional connectivity
ROI FA
DMN/FA
|
22.0, 14–81
2.961, 2.512–3.570
0.446, 0.425–0.479
6.640, 5.860–7.454
|
21.0, 16–62
2.377, 1.887–3.060
0.480, 0.464–0.493
4.953, 4.065–6.207
|
0.820
0.010*
0.001**
< 0.001**
|
Data are given as median and range except for sex (women/men) and shunt responsiveness (responder/non-responder). Mann-Whitney U-test was used except for sex and shunt responsiveness (the chi-square test).
DMN: default mode network; DMN/FA: DMN/FA: the default mode network connectivity/the mean fractional anisotropy value within the brain white matter region of interest; DR/FAB: the delayed recall score of Rey auditory verbal learning test/the FAB score; FAB: the frontal assessment battery; HCs: healthy controls; iNPH: idiopathic normal pressure hydrocephalus; iNPHGS: the idiopathic Normal Pressure Hydrocephalus Grading Scale; IR/FAB: the immediate recall score of Rey auditory verbal learning test/the FAB score; MMSE: the Mini-Mental State Examination; RAVLT: the Rey auditory verbal learning test; Re/FAB: the recognition score of Rey auditory verbal learning test/the FAB score; ROI FA: the mean fractional anisotropy value within the brain white matter region of interest; TUG: the Timed Up and Go Test.
*p < 0.05. **p < 0.005.
Table 4 and Supplementary Table 1 show the change in clinical measures after shunt placement in the preserved and reduced DMN connectivity groups. In the results of the Mann-Whitney U test, the FAB and RAVLT immediate recall scores in the preserved DMN connectivity group were significantly more improved than those in the reduced DMN connectivity group. On the other hand, there were no significant differences in variation in other clinical measures we examined between the two groups. The Wilcoxon signed rank test revealed significant improvements after shunt placement in all clinical measures we examined except for the iNPHGS cognition score in the preserved DMN connectivity group. In contrast, there were no significant improvements after shunt placement in any of the clinical measures we examined in the reduced DMN connectivity group.
Table 4
Variation of clinical measures after shunt placement in patients with iNPH.
Variables
|
Preserved DMN connectivity group
(n = 12)
|
Reduced DMN connectivity group
(n = 6)
|
P-value
|
ΔiNPHGS
Gait
Cognition
Urination
Total
ΔMMSE
ΔFAB
ΔRAVLT
Immediate recall
Delayed recall
Recognition
ΔTUG
Time to complete (second)
Number of steps
|
-1.0, -1-0
0.0, -1-0
0.0, -1-0
-1.0, -2-0
2.0, 0–5
2.0, 0–4
3.0, -6-8
2.0, -1-4
1.5, -1-4
-1.4, -33.5-1.4
-2.5, -49-0
|
-0.5, -1-0
0.0, 0–0
0.0, -1-1
-0.5, -2-0
1.0, -2-9
-1.0, -2-2
-0.5, -2-2
0.5, 0–2
-0.5, -3-4
-2.8, -27.8-2.6
-4.0, -32-2
|
0.820
0.437
0.553
0.250
0.213
0.010*
0.010*
0.291
0.180
0.820
1.000
|
Δ means that a baseline value subtracted from a postoperative value. Data are given as median and range. Mann-Whitney U-test was used.
DMN: default mode network; FAB: the frontal assessment battery; HCs: healthy controls; iNPH: idiopathic normal pressure hydrocephalus; iNPHGS: the idiopathic Normal Pressure Hydrocephalus Grading Scale; MMSE: the Mini-Mental State Examination; RAVLT: the Rey auditory verbal learning test; TUG: the Timed Up and Go Test.
*p < 0.05.
The ROC curves of the immediate recall, delayed recall, and recognition scores of the RAVLT and IR/FAB and DR/FAB ratios for differentiating the preserved DMN connectivity group from the reduced DMN connectivity group were drawn. The areas under the ROC curves were as follows: RAVLT immediate recall score, 0.875; RAVLT delayed recall score, 0.799; RAVLT recognition score, 0.806; IR/FAB ratio, 0.944; and DR/FAB ratio, 0.799. The best predictor for distinguishing the preserved and reduced DMN connectivity groups was the IR/FAB ratio. The ROC curve of the IR/FAB ratio yielded the optimal cut-off value of 1.652, and based on this cut-off value, the sensitivity was 83.3%, and the specificity was 100%.