Table 5.
Multivariate binomial logistic regression analyses of predictors for dementia in pAGD and control cases
| OR | 95% CI | p | |
|---|---|---|---|
| Quantity of AGs in the amygdala | |||
| Model 1 (n = 50) | |||
| Age at death | 1.09 | 0.98–1.22 | 0.1122 |
| Braak NFT stages III-IV | 2.08 | 0.34–12.79 | 0.4280 |
| One to 49 AGs in the amygdala (per × 400 visual field) | 1.57 | 0.30–8.31 | 0.5928 |
| Presence of LATE-NC | 2.51 | 0.27–23.78 | 0.4221 |
| Model 2 (n = 50) | |||
| Age at death | 1.09 | 0.98–1.22 | 0.1159 |
| Braak NFT stages III-IV | 1.74 | 0.26–11.74 | 0.5690 |
| 50–99 AGs in the amygdala (per × 400 visual field) | 2.43 | 0.17–34.96 | 0.5139 |
| Presence of LATE-NC | 2.03 | 0.18–23.19 | 0.5689 |
| Model 3 (n = 50) | |||
| Age at death | 1.07 | 0.95–1.20 | 0.2394 |
| Braak NFT stages III-IV | 3.55 | 0.46–27.14 | 0.2224 |
| 100 or more AGs in the amygdala (per × 400 visual field) | 10.02 | 1.12–89.43 | 0.0391* |
| Presence of LATE-NC | 3.67 | 0.32–42.61 | 0.2985 |
| Quantity of AGs in the hippocampal CA1 | |||
| Model 4 (n = 51) | |||
| Age at death | 1.07 | 0.97–1.17 | 0.1669 |
| Braak NFT stages III-IV | 3.08 | 0.55–17.23 | 0.1998 |
| One to 49 AGs in the CA1 (per × 400 visual field) | 3.63 | 0.56–23.50 | 0.1754 |
| Presence of LATE-NC | 5.13 | 0.57–45.82 | 0.1432 |
| Model 5 (n = 51) | |||
| Age at death | 1.07 | 0.97–1.18 | 0.1913 |
| Braak NFT stages III-IV | 2.83 | 0.49–16.29 | 0.2441 |
| 20–99 AGs in the CA1 (per × 400 visual field) | 0.76 | 0.08–7.03 | 0.8054 |
| Presence of LATE-NC | 4.55 | 0.41–50.60 | 0.2176 |
| Model 6 (n = 51) | |||
| Age at death | 1.07 | 0.97–1.17 | 0.1704 |
| Braak NFT stages III-IV | 2.19 | 0.36–13.34 | 0.3954 |
| 100 or more AGs in the CA1 (per × 400 visual field) | 12.22 | 1.70–87.81 | 0.0128* |
| Presence of LATE-NC | 5.43 | 0.35–83.90 | 0.2258 |
| Presence of AGs in the neocortex | |||
| Model 7 (n = 49) | |||
| Age at death | 1.06 | 0.95–1.18 | 0.2686 |
| Braak NFT stages III-IV | 2.03 | 0.27–15.20 | 0.4890 |
| Presence of AGs in the lateral occipitotemporal gyrus | 5.74 | 0.97–34.07 | 0.0544 |
| Presence of LATE-NC | 2.05 | 0.18–24.03 | 0.5673 |
| Model 8 (n = 49) | |||
| Age at death | 1.06 | 0.96–1.18 | 0.2484 |
| Braak NFT stages III-IV | 1.61 | 0.19–13.75 | 0.6622 |
| Presence of AGs in the inferior temporal gyrus | 8.18 | 1.03–65.13 | 0.0471* |
| Presence of LATE-NC | 3.33 | 0.28–38.96 | 0.3377 |
| Combination of AGs in the amygdala, CA1, or neocortex | |||
| Model 9 (n = 49) | |||
| Age at death | 1.04 | 0.92–1.17 | 0.5015 |
| Braak NFT stages III-IV | 1.82 | 0.17–19.62 | 0.6225 |
| Either 100 or more AGs per × 400 visual field in the amygdala or CA1 or the presence of AGs in the inferior temporal gyrus | 13.92 | 2.12–91.49 | 0.0061** |
| Presence of LATE-NC | 5.72 | 0.39–83.79 | 0.2025 |
51 cases (23 pAGD cases and 28 control cases) that lacked two or more lacunae or larger infarctions in the cortex and/or subcortical regions were examined using multivariate analyses.
pAGD pure argyrophilic grain disease, OR odds ratio, CI confidence interval, NFT neurofibrillary tangle, AG argyrophilic grains, LATE-NC limbic-predominant age-related TDP-43 encephalopathy neuropathologic change.
*: p < 0.05, **: p < 0.01