Data-driven FDG-PET subtypes of Alzheimer’s disease-related neurodegeneration

Table 2 Demographic, clinical and biomarker characteristics of prodromal AD subtypes at baseline

	Prodromal AD group, no hypometabolism (S0)	Prodromal AD group, limbic-predominant (S1)	Prodromal AD group, typical (S2)	P value, global comparison (S0, S1 and S2)	Pair-wise comparisons
	Prodromal AD group, no hypometabolism (S0)	Prodromal AD group, limbic-predominant (S1)	Prodromal AD group, typical (S2)	P value, global comparison (S0, S1 and S2)	S0 vs S1	S0 vs S2	S1 vs S2
Demographics
n (%)	57 (26.3%)	108 (49.8%)	49 (22.6%)
Age, years	68.4 (6.6)	76.1 (5.7)	71.7 (6.2)	< 0.001	< 0.001	0.009	< 0.001
Sex, female (%)	60%	35%	41%	0.01	0.024	0.226	1
Education, years	16.2 (2.8)	15.7 (3.0)	16.4 (2.6)	0.344
Cognition
MMSE	28.2 (1.8)	27.6 (1.8)	27.3 (1.8)	0.053
ADNI-MEM	0.57 (0.63)	0.05 (0.61)	− 0.06 (0.65)	< 0.001	< 0.001	< 0.001	0.602
ADNI-EF	0.76 (0.91)	0.00 (0.77)	0.15 (1.01)	0.018	< 0.001	0.001	0.542
ADNI-DIFF	− 0.19 (0.83)	0.05 (0.78)	− 0.21 (0.77)	0.616
ADNI-VS	0.08 (0.65)	− 0.06 (0.76)	− 0.21 (0.67)	0.135
ADNI-Lan	0.6 (0.7)	− 0.05 (0.74)	0.29 (0.79)	0.007	< 0.001	0.062	0.025
Biomarkers
APOE ε4 (%)	68%	62%	77%	0.163
AV45-PET SUVR	1.31 (0.17)	1.39 (0.17)	1.43 (0.15)	0.004	0.015	0.002	0.238
CSF Aβ, pg/ml	921 (437)	736 (237)	672 (214)	0.002	< 0.001	< 0.001	0.417
CSF t-tau, pg/ml	315 (134)	337 (137)	357 (144)	0.123
CSF p-tau, pg/ml	31 (15)	34 (16)	37 (16)	0.096
HV	4.86 (0.45)	4.53 (0.51)	4.57 (0.41)	0.009	< 0.001	0.005	1
CTV	89.87 (5.45)	83.19 (6.48)	82.93 (5.35)	< 0.001	< 0.001	< 0.001	1
HV:CTV ratio	54.26 (6.07)	54.57 (6.27)	55.24 (5.56)	0.344
WMH	4.9 (5.0)	11.3 (12.8)	6.8 (6.5)	0.172

Cortical-predominant subtype of prodromal AD group (n = 3) not included. Values for variables are presented as percentages (for gender and APOE ε4 genotype) or means with standard deviation in parentheses. Missing values are excluded (for numbers of missing values per subtype, see Supplementary Table 8). In case of significant main effects, subtypes were compared with the post hoc pairwise t tests with FDR correction. Please note that composite cognitive scores have arbitrary units on scales centred on the original test samples used to develop these scores (including patients and healthy controls) [18,19,20]
S0 no hypometabolism subtype, S1 limbic-predominant subtype, S2 typical subtype, HV hippocampal grey matter volume scaled to total intracranial volume, CTV cortical composite grey matter volume scaled to total intracranial volume

ISSN: 1758-9193