Table 1.

Summary statistics describing the empirical distribution of clinical and demographic variables in symptom progression subtypes identified in ADNI and NACC at time of patient’s dementia diagnosis

	N	MMSE	FAQ	CDRSB	Age (years)	Education (years)	APOE ε4 positive (%)	Female (%)
ADNI clusters
Slow	106	24.21 ± 2.71	11.46 ± 5.9	3.98 ± 1.36	75.32 ± 6.95	15.91 ± 2.73	67.92	25.47
Fast	177	23.55 ± 3.33	12.79 ± 5.62	4.48 ± 1.81	75.53 ± 7.04	15.9 ± 2.88	76.84	36.16
Difference (95% CI)		−0.66 (−1.5, 0.17)	1.33 (−0.22, 2.88)	0.49 (0.05, 0.94)	0.2 (−1.68, 2.09)	0.0 (−0.76, 0.75)	−8.91 (−19.84, 1.65)	10.69 (−0.58, 21.04)
NACC clusters (predicted using the ADNI-trained model)
Slow	1078	25.02 ± 4.12	14.57 ± 12.44	3.75 ± 3.19	77.68 ± 9.12	15.64 ± 3.22	48.42	51.58
Fast	1716	23.91 ± 4.2	16.43 ± 10.31	3.88 ± 2.29	78.68 ± 9.18	15.64 ± 3.14	57.75	52.39
Difference (95% CI)		−1.11 (−1.49, −0.73)	1.86 (0.99, 2.72)	0.14 (−0.07, 0.34)	1.01 (0.31, 1.70)	0.0 (−0.24, 0.24)	−9.33 (−13.1, −5.53)	0.81 (−2.99, 4.61)

Emboldened font indicates statistical significance. Numerical variables: mean ± standard deviation; the difference between subtypes is quantified as the difference in means and its 95% CI. Categorical variables: proportion of APOE ε4 carriers and female patients, respectively; differences across subtypes are quantified as the difference in proportions and its 95% CI.