Table 2.
Retrospective analysis of NACC dataset for cognition and clinical diagnosis reversion or conversion
| Imipramine vs. other antidepressants | Olanzapine vs. other antipsychotics | |||||
|---|---|---|---|---|---|---|
| N subjects imipramine; other anti-depressants | Estimate (95% C.I.) |
P-val | N subjects olanzapine; other anti-psychotics | Estimate (95% C.I.) |
P-val | |
| Cognitive exam, ΔMMSE score/year | ||||||
| All subjects | 40; 6,233 |
0.4186 (0.0017, 0.8355) |
0.0490 | 94; 798 |
0.4937 (0.0451, 0.9423) |
0.0310 |
| APOE4 carriers | 9; 2,748 |
0.6017 (-0.3156, 1.5190) |
0.1985 | 51; 354 |
0.7781 (0.1051, 1.4512) |
0.0235 |
| APOE4 non-carriers | 31; 3,485 |
0.1303 (-0.3812, 0.6419) |
0.6175 | 43; 444 |
0.3755 (-0.3011, 1.0520) |
0.2766 |
| Clinical diagnosis reversion, hazard ratioa | ||||||
| All subjects | 22, 5,476 |
1.4487 (1.2280, 1.7092) |
<0.0001 | 74; 679 |
1.7254 (0.7131, 4.1746) |
0.2263 |
| APOE4 carriers | 7; 2,635 |
0.8667 (0.4757, 1.5790) |
0.6401 | 41; 308 |
7.0936 (1.0589, 47.5200) |
0.0444 |
| APOE4 non-carriers | 15; 2,841 |
1.5313 (1.3976, 1.6778) |
<0.0001 | 33; 371 |
1.7422 (0.2738, 11.0870) |
0.9983 |
| Clinical diagnosis conversion, hazard ratiob | ||||||
| All subjects | 23; 3,987 |
0.9352 (0.7905, 1.1064) |
0.9352 | 24; 191 |
1.3000 (0.8178, 2.0666) |
0.2672 |
| APOE4 carriers | 7; 1,482 |
0.5286 (0.2585, 1.0808) |
0.0806 | 11; 57 |
1.9395 (0.9464, 3.9748) |
0.0704 |
| APOE4 non-carriers | 16; 2,505 |
1.1177 (0.9240, 1.3521) |
0.2519 | 13; 134 |
0.7770 (0.3500, 1.7246) |
0.5350 |
| APOE4 carriers vs. non-carriers | 23; 3,987 |
0.4729 (0.2256, 0.9915) |
0.0474 | 24; 191 |
2.4963 (0.8532, 7.3033) |
0.0949 |
Cognitive exam (MMSE) scores and clinical diagnosis reversion or conversion were compared between imipramine or olanzapine and control groups using regression modeling and Cox proportional hazard ratios, respectively. Summary statistics for all analyses are provided in Additional file 14. aOnly subjects who reported use of a medication prior to reversion to a better clinical diagnosis were included. bOnly subjects who reported use of a medication prior to conversion to a worse clinical diagnosis were included