correct for whole brain volume.
BELOW: SUMMARY OF DATA BETWEEN DIABETIC AND NON-DIABETIC PTS.
10% OF NON-DIABETIC PATIENTS CONVERTED TO DEMENTIA, WHILE NEARLY 3X AS MANY (29%) DIABETIC PATIENTS CONVERTED TO DEMENTIA. CONVERSION IS A SIGNFICANT FACTOR (P = 0.027).
|
SUMMARY OF PT DATA BY DIABETES STATUS
|
| Characteristic |
No DM2, N = 150 |
Yes DM2, N = 21 |
p-value |
| convert |
15 (10%) |
6 (29%) |
0.027 |
| APOE4 |
|
|
0.6 |
| 0 |
78 (58%) |
15 (71%) |
|
| 1 |
44 (33%) |
5 (24%) |
|
| 2 |
13 (9.6%) |
1 (4.8%) |
|
| Unknown |
15 |
0 |
|
| MMSE |
28.0 (26.0, 30.0) |
29.0 (27.0, 30.0) |
0.088 |
| Unknown |
24 |
0 |
|
BELOW: SUMMARY OF DATA BETWEEN DEMENTIA-CONVERTED (1; right column) AND NON-CONVERTED (0; left column) PTS.
CONVERTED PATIENTS HAD A SIGNIFICANTLY LOWER MMSE (24.0 vs 29.0). CONVERTED PATIENTS WERE PROPORTIONALLY HIGHER DIABETIC (29% CONVERTED vs 10% UNCONVERTED).
|
SUMMARY OF PT DATA BY DEMENTIA CONVERSION
|
| Characteristic |
0, N = 150 |
1, N = 21 |
p-value |
| APOE4 |
|
|
0.9 |
| 0 |
80 (59%) |
13 (62%) |
|
| 1 |
42 (31%) |
7 (33%) |
|
| 2 |
13 (9.6%) |
1 (4.8%) |
|
| Unknown |
15 |
0 |
|
| MMSE |
29.0 (27.0, 30.0) |
24.0 (19.0, 27.0) |
<0.001 |
| Unknown |
24 |
0 |
|
| dm2 |
|
|
0.027 |
| No DM2 |
135 (90%) |
15 (71%) |
|
| Yes DM2 |
15 (10%) |
6 (29%) |
|
BELOW: LOG MODEL PREDICTS DEMENTIA CONVERSION (1 CONVERTED / 0 NOT ) BY APOE4, GENDER, AGE, MMSE, AND DM2 STATUS.
GENDER, MMSE, AND DIABETES STATUS ARE SIGNFNICANT FACTORS. DIABETES STATUS APPEARS TO HAVE THE STRONGEST RELATIONSHIP WITH CONVERSION.
|
Observations
|
132 (39 missing obs. deleted)
|
|
Dependent variable
|
convert
|
|
Type
|
Generalized linear model
|
|
Family
|
binomial
|
|
Link
|
logit
|
|
𝛘²(5)
|
42.44
|
|
Pseudo-R² (Cragg-Uhler)
|
0.47
|
|
Pseudo-R² (McFadden)
|
0.37
|
|
AIC
|
85.23
|
|
BIC
|
102.53
|
|
|
Est.
|
S.E.
|
z val.
|
p
|
|
(Intercept)
|
7.41
|
4.82
|
1.54
|
0.12
|
|
APOE4
|
0.13
|
0.48
|
0.27
|
0.79
|
|
PTGENDERMale
|
1.71
|
0.78
|
2.20
|
0.03
|
|
AGE
|
0.03
|
0.05
|
0.70
|
0.49
|
|
MMSE
|
-0.52
|
0.11
|
-4.48
|
0.00
|
|
dm2_bin
|
2.78
|
0.87
|
3.19
|
0.00
|
|
Standard errors: MLE
|
##
## Call:
## glm(formula = convert ~ APOE4 + PTGENDER + AGE + MMSE + dm2_bin,
## family = binomial, data = dfx)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.09461 -0.49563 -0.23252 -0.09763 2.85383
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 7.41262 4.82104 1.538 0.12416
## APOE4 0.12973 0.47590 0.273 0.78516
## PTGENDERMale 1.71165 0.77879 2.198 0.02796 *
## AGE 0.03385 0.04869 0.695 0.48693
## MMSE -0.51536 0.11491 -4.485 7.29e-06 ***
## dm2_bin 2.78361 0.87262 3.190 0.00142 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 115.674 on 131 degrees of freedom
## Residual deviance: 73.232 on 126 degrees of freedom
## (39 observations deleted due to missingness)
## AIC: 85.232
##
## Number of Fisher Scoring iterations: 6
BELOW: LOG MODEL PREDICTS DIABETES STATUS (1 DIABETIC / 0 NOT ) BY APOE4, GENDER, AGE, MMSE, DEMENTIA CONVERSION AND DM2 STATUS.
AGE, MMSE, AND DEMENTIA CONVERSION ARE SIGNFNICANT FACTORS. DEMENTIA CONVERSION APPEARS TO HAVE THE STRONGEST RELATIONSHIP WITH DIABETES STATUS.
|
Observations
|
132 (39 missing obs. deleted)
|
|
Dependent variable
|
dm2_bin
|
|
Type
|
Generalized linear model
|
|
Family
|
binomial
|
|
Link
|
logit
|
|
𝛘²(5)
|
19.99
|
|
Pseudo-R² (Cragg-Uhler)
|
0.24
|
|
Pseudo-R² (McFadden)
|
0.17
|
|
AIC
|
107.68
|
|
BIC
|
124.98
|
|
|
Est.
|
S.E.
|
z val.
|
p
|
|
(Intercept)
|
-4.19
|
4.92
|
-0.85
|
0.39
|
|
APOE4
|
-0.43
|
0.43
|
-1.01
|
0.31
|
|
PTGENDERMale
|
-0.27
|
0.58
|
-0.47
|
0.64
|
|
AGE
|
-0.08
|
0.04
|
-1.85
|
0.06
|
|
MMSE
|
0.28
|
0.12
|
2.29
|
0.02
|
|
convert
|
2.26
|
0.81
|
2.77
|
0.01
|
|
Standard errors: MLE
|
BELOW: LIST OF BRAIN REGIONS (FDR CORRECTED) SIGNFICANTLY DIFFERENT IN DIABETES PATIENTS
|
|
Code
|
Region
|
FDR_corrected
|
FDR_sig
|
metric
|
|
11
|
ST105SA
|
Surface Area (aparc.stats) of RightPericalcarine
|
0.0315366
|
YES
|
Surface Area
|
|
21
|
ST111SA
|
Surface Area (aparc.stats) of RightRostralMiddleFrontal
|
0.0330164
|
YES
|
Surface Area
|
|
23
|
ST113SA
|
Surface Area (aparc.stats) of RightSuperiorFrontal
|
0.0394345
|
YES
|
Surface Area
|
|
27
|
ST115SA
|
Surface Area (aparc.stats) of RightSuperiorTemporal
|
0.0330164
|
YES
|
Surface Area
|
|
39
|
ST14SA
|
Surface Area (aparc.stats) of LeftCuneus
|
0.0330164
|
YES
|
Surface Area
|
|
47
|
ST25SA
|
Surface Area (aparc.stats) of LeftHemisphereWM
|
0.0330164
|
YES
|
Surface Area
|
|
84
|
ST52SA
|
Surface Area (aparc.stats) of LeftRostralMiddleFrontal
|
0.0330164
|
YES
|
Surface Area
|
|
90
|
ST56SA
|
Surface Area (aparc.stats) of LeftSuperiorTemporal
|
0.0056462
|
YES
|
Surface Area
|
|
94
|
ST58SA
|
Surface Area (aparc.stats) of LeftTransverseTemporal
|
0.0213410
|
YES
|
Surface Area
|
|
127
|
ST98SA
|
Subcortical Volume (aseg.stats) of CorticalWM
|
0.0208046
|
YES
|
Cortical
|
|
128
|
ST98TA
|
Subcortical Volume (aseg.stats) of TotalGM
|
0.0208046
|
YES
|
Cortical
|
