Table 2.
Association of KL*VSHET+ with amyloid PET positivity status, stratified by APOE*4 status, in cognitively normal subjects aged 60–80
| Study stratum | Association between KL*VSHET+ and Amy+ by APOE*4 status | Interaction between KL*VSHET+ and Amy+ by APOE*4 status | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Amy− with KL*VSHET+ (N/total) |
Amy+ with KL*VSHET+ (N/total) |
Odds ratio [95% CI] |
p-value | Power | Amy− with KL*VSHET+ (N/total) |
Amy+ with KL*VSHET+ (N/total) |
Odds ratio [95% CI] |
p-value | Power | |
| ADNI | ||||||||||
| APOE*4+ | 14/24 (58.3%) | 10/48 (20.8%) | 0.18 [0.06–0.59] | 0.0044 | 0.16 (0.11–0.21) | 47/132 (35.6%) | 25/97 (25.8%) | 0.18 [0.05–0.70] | 0.013 | 0.08 [0.05–0.14] |
| APOE*4− | 33/108 (30.6%) | 15/49 (30.6%) | 0.99 [0.66–1.32] | 0.98 | 0.14 [0.08–0.18] | |||||
| A4 | ||||||||||
| APOE*4+ | 129/450 (28.7%) | 98/426 (23.0%) | 0.72 [0.53–0.98] | 0.038 | 0.77 [0.45–0.96] | 446/1662 (26.8%) | 149/632 (23.6%) | 0.77 [0.49–1.23] | 0.28 | 0.50 [0.25–0.72] |
| APOE*4− | 317/1212 (26.2%) | 51/206 (24.8%) | 0.93 [0.66–1.32] | 0.69 | 0.62 [0.32–0.88] | |||||
| AIBL | ||||||||||
| APOE*4+ | 21/59 (35.6%) | 19/85 (22.4%) | 0.53 [0.25–1.11] | 0.092 | 0.27 [0.15–0.38] | 90/332 (27.1%) | 41/183 (22.4%) | 0.61 [0.24–1.55] | 0.30 | 0.19 [0.15–0.32] |
| APOE*4− | 69/273 (25.3%) | 22/98 (22.4%) | 0.87 [0.50–1.51] | 0.62 | 0.31 [0.19–0.54] | |||||
| Insight 46 | ||||||||||
| APOE*4+ | 17/75 (22.7%) | 9/42 (21.5%) | 0.90 [0.35–2.29] | 0.82 | 0.18 [0.10–0.26] | 85/342 (24.9%) | 18/73 (24.7%) | 0.80 [0.23–2.77] | 0.73 | 0.08 [0.05–0.10] |
| APOE*4− | 68/267 (25.5%) | 9/31 (26.7%) | 1.33 [0.57–3.07] | 0.51 | 0.02 [0.02–0.02] | |||||
| HABS | ||||||||||
| APOE*4+ | 1/13 (7.7%) | 9/30 (30.0%) | 6.09 [0.56–66.5] | 0.14 | 0.04 [0.04–0.06] | 18/79 (22.8%) | 13/49 (26.5%) | 7.47 [0.54–102.6] | 0.13 | 0.03 [0.02–0.03] |
| AP0E*4− | 17/66 (25.8%) | 4/19 (21.1%) | 0.56 [0.13–2.37] | 0.43 | 0.021 [0.02–0.02] | |||||
| Meta-analysisa | ||||||||||
| APOE*4+ | 182/621 (29.3%) | 145/631 (23.0%) | 0.67 [0.52–0.88] | 0.0035 | 0.90 [0.57–0.99] | 686/2547 (26.9%) | 246/1034 (23.8%) | 0.70 [0.48–1.021] | 0.061 | 0.65 [10.33–0.88] |
| APOE*4− | 504/1926 (26.2%) | 101/403 (25.1%) | 0.94 [0.73–1.21] | 0.63 | 0.86 [0.52–0.99] | |||||
| Meta-analysis without ADNIb | ||||||||||
| APOE*4+ | 168/597 (28.1%) | 135/583 (23.2%) | 0.72 [0.55–0.95] | 0.020 | 0.85 [0.54–0.99] | 639/2415 (26.5%) | 221/937 (23.6%) | 0.78 [0.53–1.16] | 0.22 | 0.60 [0.34–0.86] |
| APOE*4− | 471/1818 (25.9%) | 86/354 (24.3%) | 0.93 [0.71–1.23] | 0.62 | 0.82 [0.49–0.98] | |||||
Power is directly reported in the table for an OR of 0.7 and additionally for OR values ranging from 0.6 to 0.8 [denoted by square brackets], corresponding to a priori expected effect sizes (cf. methods).
Key: Amy+, amyloid positive; Amy−, amyloid negative; HET+, heterozygous carriers; OR, odds ratio; CI, confidence interval.
Cochran’s Q tests for heterogeneity were non-significant for the displayed meta-analyses across all cohorts in the APOE*4+ (Q = 9.01, p = 0.06), APOE*4− (Q = 1.24, p = 0.87), and full sample (Q = 7.24, p = 0.12).
Meta-analyses were repeated after excluding ADNI to ensure a fully independent validation effort of prior work (Belloy et al., 2020). Cochran’s Q tests for heterogeneity were non-significant for the displayed meta-analyses across cohorts, when excluding ADNI, in the APOE*4+ (Q = 3.95, p = 0.27), APOE*4− (Q = 1.22, p = 0.75), and full sample (Q = 3.12, p = 0.37).