| |
Cognitively normal
|
MCI
|
AD
|
|---|
|
(n = 37)
|
(n = 61)
|
(n = 65)
|
|---|
|
YKL-40 (ng/ml), baseline
|
231 (16)
|
304 (16)*
|
288 (12)*
|
|
YKL-40 (ng/ml), follow-up
|
241 (18)
|
320 (16)*
|
306 (14)†
|
|
Annual change, (β (SE))
|
5.3 (3.2)
|
8.9 (3.0)‡
|
7.1 (3.1)§
|
|
VILIP-1 (pg/ml), baseline
|
168 (11)
|
192 (13)
|
182 (10)
|
|
VILIP-1 (pg/ml), follow-up
|
175 (11)
|
217 (14)
|
190 (11)
|
|
Annual change, (β (SE))
|
2.8 (2.8)
|
10.7 (2.6)‡
|
3.1 (2.6)
|
- Data presented as mean (SE). At baseline, VILIP-1 data were missing for two patients and YKL-40 data for one patient; at follow-up, VILIP-1 data were missing for one patient and YKL-40 data for three patients. Baseline and follow-up differences (mean (SE) LP interval was 2.0 (0.1) years) were assessed with ANOVA with post-hoc Bonferroni corrections, adjusted for age and sex. CSF biomarkers were log-transformed for ANOVA analyses. Longitudinal effects were assessed using age and sex-adjusted linear mixed models, with CSF biomarkers (VILIP-1 and YKL-40) as dependent variables and clinical diagnosis (categorical), time (LP interval in years), and interaction diagnosis × time as independent variables. The reported β value represents the estimated change of YKL-40 (ng/ml) or VILIP-1 (pg/ml) levels per year
-
*
p ≤0.05 vs. cognitively normal subjects
-
†
p ≤0.005 vs. cognitively normal subjects
-
‡
p ≤0.005 for time effect
-
§
p ≤0.05 for time effect
-
AD Alzheimer’s disease, ANOVA analysis of variance, CSF cerebrospinal fluid, LP lumbar puncture, MCI mild cognitive impairment, SE standard error, VILIP-1 Visinin-like protein-1, YKL-40 Chitinase-3-like protein 1
