Table 3.
Lag time analyses on comparative risk of dementia in main propensity score matched cohort, with follow-up starting after 365 days from index date
| No of propensity score matched pairs | SGLT-2 inhibitors | DPP-4 inhibitors (ref) | Hazard ratio (95% CI) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Events | Person years | Incidence rate per 100 person years (95% CI) | Events | Person years | Incidence rate per 100 person years (95% CI) | Cox model | Fine-Gray model | ||||
| As treated analysis | |||||||||||
| Dementia | 34 048 | 159 | 69 466 | 0.23 (0.19 to 0.26) | 293 | 77 010 | 0.38 (0.34 to 0.42) | 0.57 (0.48 to 0.68) | 0.61 (0.50 to 0.74) | ||
| Dementia requiring drugs | 34 113 | 89 | 69 665 | 0.13 (0.10 to 0.15) | 191 | 77 305 | 0.25 (0.21 to 0.28) | 0.48 (0.38 to 0.61) | 0.53 (0.41 to 0.68) | ||
| Alzheimer’s disease | 34 085 | 129 | 69 569 | 0.19 (0.15 to 0.22) | 246 | 77 198 | 0.32 (0.28 to 0.36) | 0.55 (0.45 to 0.67) | 0.59 (0.48 to 0.73) | ||
| Vascular dementia | 34 180 | 13 | 69 882 | 0.02 (0.01 to 0.03) | 33 | 77 634 | 0.04 (0.03 to 0.06) | 0.46 (0.26 to 0.80) | 0.44 (0.23 to 0.84) | ||
| Genital infection | 31 048 | 1634 | 60 245 | 2.71 (2.58 to 2.84) | 995 | 68 083 | 1.46 (1.37 to 1.55) | 1.92 (1.79 to 2.06) | 1.81 (1.68 to 1.96) | ||
| Osteoarthritis related encounters | 21 991 | 3900 | 37 979 | 10.3 (10.0 to 10.6) | 4437 | 41 468 | 10.7 (10.4 to 11.0) | 0.95 (0.91 to 0.99) | 0.96 (0.92 to 1.00) | ||
| Cataract surgery | 32 061 | 1823 | 63 009 | 2.89 (2.76 to 3.03) | 2253 | 69 216 | 3.26 (3.12 to 3.39) | 0.93 (0.88 to 0.98) | 0.92 (0.85 to 0.96) | ||
| Intention-to-treat analysis to 3 years* | |||||||||||
| Dementia | 77 396 | 607 | 172 767 | 0.35 (0.32 to 0.38) | 770 | 173 238 | 0.44 (0.41 to 0.48) | 0.77 (0.71 to 0.84) | 0.79 (0.71 to 0.88) | ||
| Dementia requiring drugs | 77 526 | 412 | 173 307 | 0.24 (0.22 to 0.26) | 502 | 173 885 | 0.29 (0.26 to 0.31) | 0.81 (0.73 to 0.89) | 0.83 (0.73 to 0.94) | ||
| Alzheimer’s disease | 77 477 | 493 | 173 110 | 0.29 (0.26 to 0.31) | 635 | 173 639 | 0.37 (0.34 to 0.39) | 0.76 (0.69 to 0.83) | 0.78 (0.69 to 0.88) | ||
| Vascular dementia | 77 674 | 77 | 174 010 | 0.04 (0.03 to 0.05) | 96 | 174 664 | 0.06 (0.04 to 0.07) | 0.83 (0.67 to 1.03) | 0.81 (0.60 to 1.09) | ||
| Genital infection | 71 479 | 3735 | 153 363 | 2.44 (2.36 to 2.51) | 2635 | 156 056 | 1.69 (1.62 to 1.75) | 1.43 (1.37 to 1.48) | 1.44 (1.37 to 1.52) | ||
| Osteoarthritis related encounters | 53 768 | 10 843 | 103 926 | 10.4 (10.2 to 10.6) | 10 678 | 104 666 | 10.2 (10.0 to 10.4) | 1.03 (1.01 to 1.05) | 1.03 (1.00 to 1.05) | ||
| Cataract surgery | 73 730 | 4965 | 159 061 | 3.12 (3.04 to 3.21) | 5038 | 159 482 | 3.16 (3.07 to 3.25) | 0.98 (0.95 to 1.01) | 0.99 (0.95 to 1.03) | ||
| Intention-to-treat analysis to maximum follow-up* | |||||||||||
| Dementia | 77 396 | 908 | 227 961 | 0.40 (0.37 to 0.42) | 1134 | 230 043 | 0.49 (0.46 to 0.52) | 0.80 (0.75 to 0.86) | 0.82 (0.75 to 0.89) | ||
| Dementia requiring drugs | 77 526 | 632 | 228 852 | 0.28 (0.26 to 0.30) | 768 | 231 156 | 0.33 (0.31 to 0.36) | 0.84 (0.77 to 0.91) | 0.84 (0.76 to 0.93) | ||
| Alzheimer’s disease | 77 477 | 755 | 228 538 | 0.33 (0.31 to 0.35) | 949 | 230 711 | 0.41 (0.39 to 0.44) | 0.80 (0.74 to 0.86) | 0.81 (0.74 to 0.89) | ||
| Vascular dementia | 77 674 | 97 | 230 144 | 0.04 (0.03 to 0.05) | 128 | 232 555 | 0.06 (0.05 to 0.07) | 0.80 (0.66 to 0.98) | 0.76 (0.59 to 0.99) | ||
| Genital infection | 71 479 | 4336 | 199 630 | 2.17 (2.11 to 2.24) | 3192 | 205 272 | 1.56 (1.50 to 1.61) | 1.38 (1.34 to 1.43) | 1.39 (1.33 to 1.46) | ||
| Osteoarthritis related encounters | 53 768 | 12 965 | 130 000 | 9.97 (9.80 to 10.15) | 12 747 | 131 984 | 9.66 (9.49 to 9.83) | 1.03 (1.01 to 1.05) | 1.03 (1.01 to 1.06) | ||
| Cataract surgery | 73 730 | 6678 | 206 281 | 3.24 (3.16 to 3.32) | 6810 | 207 999 | 3.27 (3.20 to 3.35) | 1.00 (0.97 to 1.02) | 0.99 (0.96 to 1.03) | ||
CI=confidence interval; DPP-4=dipeptidyl peptidase-4; SGLT-2=sodium glucose co-transporter 2.
*
One year lag time applied in intention-to-treat analysis showed attenuated association. Because patients remained in the index treatment group even if they discontinued or switched from their index treatment, misclassification of drug use is least for the initial follow-up period. Therefore, starting follow-up one year after the index date will result in greater misclassification of drug use and drive the effect estimate towards null in the intention-to-treat analysis.