Baseline characteristics of patients initiating DPP-4 inhibitors, SGLT2 inhibitors or GLP-1 receptor agonists
| Characteristic | DPP-4i | SGLT2i | GLP-1RA | Total cohort |
| No of patients | 3817 (50.7) | 2851 (37.9) | 855 (11.4) | 7523 (100) |
| Gender (female) (n, %)* | 1849 (48.4) | 1088 (38.2) | 392 (45.8) | 3329 (44.3) |
| Patient age (mean, SD)* | 69.2 (11.0) | 64.0 (8.4) | 61.7 (8.8) | 66.4 (10.3) |
| Patient age (median, IQR) | 69 (61–78) | 64 (58–70) | 62 (56–68) | 66 (59–74) |
| Duration of diabetes therapy* | ||||
| Less than 5 years | 738 (19.3) | 374 (13.1) | 139 (16.3) | 1252 (16.6) |
| 5 years or more | 3079 (80.7) | 2477 (86.9) | 716 (83.7) | 6271 (83.4) |
| Antidiabetic medicines used in the past 135 days | ||||
| Metformin* | 3009 (78.8) | 2626 (92.1) | 790 (92.4) | 6425 (85.4) |
| Sulphonylureas* | 3029 (79.4) | 1970 (69.1) | 577 (67.5) | 5576 (74.1) |
| Other antidiabetic medicines†* | 134 (3.5) | 73 (2.6) | 39 (4.6) | 246 (3.3) |
| Number of different antidiabetic medicine classes used in the past 135 days‡ | ||||
| One antidiabetic medicine class | 1473 (38.6) | 1048 (36.8) | 310 (36.3) | 2831 (37.6) |
| Two antidiabetic medicine classes | 2333 (61.1) | 1787 (62.7) | 539 (63.0) | 4659 (61.9) |
| Three antidiabetic medicine classes | 11 (0.3) | 16 (0.6) | 6 (0.7) | 33 (0.4) |
| Use of insulin | ||||
| Use of insulin therapy in the past year* | 248 (6.5) | 798 (28.0) | 333 (38.9) | 1379 (18.3) |
| Previous hospitalisation | ||||
| Hospital admission due to CV causes in the past year* | 263 (6.9) | 171 (6.0) | 42 (4.9) | 476 (6.3) |
| Hospital admission due to CV causes (excluding HF) in the past year | 224 (5.9) | 160 (5.6) | 38 (4.4) | 422 (5.6) |
| Hospital admission due to HF in the past year* | 44 (1.2) | 13 (0.5) | 6 (0.7) | 63 (0.8) |
| Hospital admission due to MI or stroke in the past year | 46 (1.2) | 40 (1.4) | 6 (0.7) | 92 (1.2) |
| Hospital admission due to type 2 diabetes in the past year* | 59 (1.5) | 33 (1.2) | 27 (3.2) | 119 (1.6) |
| Hospital admission due to/with cancer in the past 5 years§* | 198 (5.2) | 116 (4.1) | 32 (3.7) | 346 (4.6) |
| Concomitant therapy (other medicines used in the past 135 days) | ||||
| Anticoagulant* | 432 (11.3) | 201 (7.1) | 54 (6.3) | 687 (9.1) |
| Platelet inhibitor* | 1380 (36.2) | 1141 (40.0) | 293 (34.3) | 2814 (37.4) |
| Antiarrhythmic, cardiac glycoside or vasodilator* | 399 (10.5) | 227 (8.0) | 68 (8.0) | 694 (9.2) |
| Loop diuretic* | 536 (14.0) | 258 (9.0) | 113 (13.2) | 907 (12.1) |
| Thiazide and other diuretic* | 1536 (40.2) | 1160 (40.7) | 391 (45.7) | 3087 (41.0) |
| MRA | 172 (4.5) | 118 (4.1) | 41 (4.8) | 331 (4.4) |
| Beta blocker* | 1489 (39.0) | 1181 (41.4) | 371 (43.4) | 3041 (40.4) |
| Calcium channel blocker* | 1423 (37.3) | 1045 (36.7) | 353 (41.3) | 2821 (37.5) |
| ACE-inhibitor or ARB¶* | 2773 (72.6) | 2164 (75.9) | 671 (78.5) | 5608 (74.5) |
| Statin* | 2211 (57.9) | 1865 (65.4) | 511 (59.8) | 4587 (61.0) |
| Other lipid modifying agent* | 183 (4.8) | 195 (6.8) | 68 (8.0) | 446 (5.9) |
| Antidepressant* | 487 (12.8) | 333 (11.7) | 141 (16.5) | 961 (12.8) |
| Anxiolytic, hypnotic, or sedative* | 587 (15.4) | 351 (12.3) | 96 (11.2) | 1034 (13.7) |
| NSAID* | 761 (19.9) | 657 (23.0) | 213 (24.9) | 1631 (21.7) |
| Time of cohort entry* | ||||
| From 30 June 2014 to 29 June 2016 | 2236 (58.6) | 837 (29.4) | 345 (40.4) | 3418 (45.4) |
| From 30 June 2016 to 30 June 2018 | 1581 (41.4) | 2014 (70.6) | 510 (59.6) | 4105 (54.6) |
*Indicates a statistically significant difference (p value for omnibus test <0.05) among antidiabetic medicine groups.
†Repaglinide or acarbose.
‡Excluding insulin.
§Cancer as a main or concomitant diagnosis; this covariate was included only in the all-cause death model as cancer was the second most common cause of death in our cohort.
¶Including combination of valsartan and sacubitril (Eight patients received combination of valsartan and sacubitril within 135 days preceding index date. Due to the small number of patients using a combination of valsartan and sacubitril, this covariate was not separately included in the Cox regression models).
ACE, angiotensin-converting enzyme inhibitors; ARB, angiotensin II receptor blockers; CV, cardiovascular; DPP-4i, dipeptidyl peptidase-4 inhibitor; GLP-1RA, glucagon-like peptide-1 receptor agonist; HF, heart failure; MI, myocardial infarction; MRA, mineralocorticoid (aldosterone) receptor antagonists; NSAID, non-steroidal anti-inflammatory drug; SGLT2i, sodium-glucose co-transporter 2 inhibitor.