Table 3.
The effect of time-varying cardiometabolic measurements on sex- and APOE ε4-related cerebral perfusion decline with age (n = 880).a
| Gray mattera | Hippocampusa | Superior frontal gyrusa | Middle frontal gyrusa | Posterior cingulatea | Precuneusa | |
|---|---|---|---|---|---|---|
| Model 1b | ||||||
| Men ε4 carriers | Ref (0.00) | Ref (0.00) | Ref (0.00) | Ref (0.00) | Ref (0.00) | Ref (0.00) |
| ε4 non-carriers | −0.83 (−1.78, 0.12) | −0.68 (−1.72, 0.35) | −1.10 (−2.20, 0.01) | −1.31 (−2.52, −0.10)* | −1.20 (−3.05, 0.65) | −1.40 (−2.84, 0.05) |
| Women ε4 carriers | −1.14 (−2.20, −0.09)* | −0.94 (−2.08, 0.21) | −1.57 (−2.79, −0.34)* | −1.82 (−3.15, −0.48)** | −1.77 (−3.81, 0.27) | −1.76 (−3.35, −0.16)* |
| Model 2 | ||||||
| Model 1+time-varying covariates of cardiometabolic measurementsc | ||||||
| Men ε4 carriers | Ref (0.00) | Ref (0.00) | Ref (0.00) | Ref (0.00) | Ref (0.00) | Ref (0.00) |
| ε4 non-carriers | −0.79 (−1.80, 0.23) | −0.69 (−0.42, 8.72) | −0.84 (−2.06, 0.39) | −1.04 (−2.37, 0.29) | −1.28 (−3.24, 0.68) | −1.43 (−3.00, 0.15) |
| Women ε4 carriers | −0.94 (−2.07, 0.19) | −0.85 (−2.06, 0.36) | −1.30 (−2.67, 0.07) | −1.51 (−2.99, −0.02)* | −1.51 (−3.70, 0.68) | −1.63 (−3.39, 0.13) |
| Likelihood-ratio test between Model 1 and Model 2 | ||||||
| Chi2 (p-value) | 61.52 (<0.001) | 40.28 (<0.001) | 64.18 (<0.001) | 72.34 (<0.001) | 46.84 (<0.001) | 56.10 (<0.001) |
| AIC | ||||||
| Model 1 | 8710.77 | 8911.84 | 9210.33 | 9404.62 | 10336.31 | 9807.97 |
| Model 2 | 8660.52 | 8879.62 | 9157.72 | 9348.28 | 10298.4 | 9763.27 |
aThere were 70 participants with missing information on APOE ε4 status.
bThe β-coefficients and 95% confidence intervals in the models were adjusted birth cohort, sex, APOE ε4 status, education, parental history of dementia, smoking status, intracranial volume, post-labeling delay, and head coil.
cTime-varying covariates were systolic blood pressure, body mass index, blood glucose, and total cholesterol.
*0.01<p < 0.05; **p < 0.01.