TABLE 3.

Efficacy of the elevated plasma Aβ1-42/Aβ1-40 ratio in diagnosing MCI in both training and validation cohorts of T2DM patients.

	Quartile of the plasma Aβ1-42/Aβ1-40
	Q1 (Lowest)	Q2	Q3	Q4 (Highest)	Increase 1 value of Aβ1-42/Aβ1-40	p trend*
Training cohort
Aβ1-42/Aβ1-40 levels	<0.247	0.247–0.434	0.434–0.754	≥0.754
Cases/controls, n	9/69	20/69	26/67	34/69
Aβ1-42/Aβ1-40, OR	1.00 (ref.)	2.22 (0.95–5.22)	2.98 (1.30–6.82)	3.78 (1.69–8.47)	2.14 (1.45–3.15)	0.002
Aβ1-42/Aβ1-40-Adjusted-1, OR	1.00 (ref.)	2.32 (0.95–5.65)	2.80 (1.18–6.67)	3.53 (1.52–8.19)	2.13 (1.39–3.28)	0.007
Aβ1-42/Aβ1-40-Adjusted-2, OR	1.00 (ref.)	3.14 (1.01–9.79)	5.33 (1.69–16.78)	6.52 (2.16–19.69)	3.15 (1.80–5.52)	0.001
Validation cohort
Aβ1-42/Aβ1-40 levels	<0.203	0.203–0.320	0.320–0.497	≥0.497
Cases/controls, n	42/56	51/56	62/57	109/56
Aβ1-42/Aβ1-40, OR	1.00 (ref.)	1.21 (0.70–2.11)	1.45 (0.85–2.48)	2.60 (1.55–4.34)	2.22 (1.56–3.15)	<0.001
Aβ1-42/Aβ1-40-Adjusted-1, OR	1.00 (ref.)	1.02 (0.57–1.82)	1.64 (0.93–2.89)	3.15 (1.83–5.43)	2.64 (1.80–3.89)	<0.001
Aβ1-42/Aβ1-40-Adjusted-2, OR	1.00 (ref.)	1.09 (0.55–2.17)	1.15 (0.58–2.28)	2.50 (1.31–4.74)	2.40 (1.62–3.55)	<0.001

OR, odds ratio (95% CI).

Aβ1-42/Aβ1-40-Adjusted-1 means the Aβ ratio was adjusted for age (in continuous), sex, hypertension (yes/no), hyperlipidemia (yes/no), CHD (yes/no), family history of diabetes (yes/no), diabetes duration (in continuous and quadratic terms), HbA1c (in continuous and quadratic terms), insulin treatment (yes/no), and diabetic complications (yes/no).

Aβ1-42/Aβ1-40-Adjusted-2 means the Aβ ratio was adjusted for Model 1, plus rGSK-3β (Total/S9) (in continuous and quadratic terms), ApoE ε4 (yes/no), and olfactory score (in continuous).

*Linear trend test using the median value of each category.