Table 4.
Results of two-variable temporal effect analyses.
| Model | Subgroup | Estimate | Std. Error | p value | 95% CI |
|---|---|---|---|---|---|
| Sex × DG group | Sex (female)* | 11.21 | 3.95 | 0.005 | (3.48, 18.94) |
| Convert* | 14.75 | 5.72 | 0.01 | (3.55, 25.97) | |
| AD | 12.87 | 11.35 | 0.257 | (−9.37, 35.07) | |
| MCI | 5.47 | 4.44 | 0.218 | (−3.23, 14.17) | |
| Sex × ApoE | Sex (female)* | 11.72 | 3.92 | 0.003 | (4.04, 19.4) |
| ApoE E4 heterozygote | −0.98 | 4.24 | 0.817 | (−9.28, 7.32) | |
| ApoE E4 homozygote* | 28.68 | 7.11 | <0.001 | (14.77, 42.6) | |
| Sex × Age | Sex (female)* | 10.44 | 3.94 | 0.008 | (2.71, 18.15) |
| Age* | −0.55 | 0.27 | 0.045 | (−1.08, −0.01) | |
| DG group × ApoE | Convert | 10.19 | 5.98 | 0.089 | (−1.52, 21.91) |
| AD | 5.65 | 11.59 | 0.626 | (−17.07, 28.33) | |
| MCI | 2.88 | 4.47 | 0.519 | (−5.87, 11.63) | |
| ApoE E4 heterozygote | −1.64 | 4.34 | 0.705 | (−10.14, 6.85) | |
| ApoE E4 homozygote* | 24.53 | 7.48 | 0.001 | (9.89, 39.16) | |
| DG group × Age | Convert* | 14.83 | 5.73 | 0.01 | (3.62, 26.05) |
| AD | 12.92 | 11.37 | 0.256 | (−9.36, 35.16) | |
| MCI | 3.53 | 4.45 | 0.427 | (−5.18, 12.24) | |
| Age* | −0.63 | 0.28 | 0.021 | (−1.17, −0.1) | |
| ApoE × Age | ApoE E4 heterozygote | −1.06 | 4.26 | 0.804 | (−9.41, 7.29) |
| ApoE E4 homozygote* | 26.41 | 7.18 | <0.001 | (12.34, 40.48) | |
| Age | −0.49 | 0.28 | 0.076 | (−1.03, 0.05) | |
| GDS × DG group | Convert* | 12.6 | 5.9 | 0.033 | (1.07, 24.15) |
| AD | 9.4 | 11.41 | 0.41 | (−12.97, 31.72) | |
| MCI | 2.72 | 4.57 | 0.552 | (−6.23, 11.66) | |
| GDS | 2.16 | 1.39 | 0.119 | (−0.55, 4.88) | |
| GDS × Sex | Sex (female)* | 10.74 | 3.93 | 0.006 | (3.03, 18.43) |
| GDS* | 2.65 | 1.33 | 0.047 | (0.04, 5.26) | |
| GDS × ApoE | ApoE E4 heterozygote | 0.21 | 4.24 | 0.96 | (−8.1, 8.52) |
| ApoE E4 homozygote* | 27.88 | 7.12 | <0.001 | (13.94, 41.82) | |
| GDS* | 2.62 | 1.33 | 0.049 | (0.02, 5.23) | |
| GDS × Age | Age | −0.52 | 0.28 | 0.06 | (−1.06, 0.02) |
| GDS | 2.36 | 1.35 | 0.081 | (−0.29, 5.01) |
Temporal effects (as obtained by time × variable interaction) only, each variable × time interaction calculated separately in the same model. Baseline corrected for sex, diagnostic group, ApoE and baseline age as appropriate. Asterisk (*) denotes significant temporal effects (even with inclusion of the second variable’s temporal effect into the model). Sex coded as 0 = males, 1 = females. For multi-level variables Dg. Group and ApoE, the comparison variable is CN and ApoE E4 non-carrier, respectively. Rounding for 2 decimal places (3 for p-value). CI, confidence interval; AD, Alzheimer’s disease dementia; MCI, mild cognitive impairment; GDS, geriatric depression scale.