Abstract
Aims
In this first interim analysis of the SCORE study, we investigated the risk of major adverse cardiovascular events (MACE) among individuals with atherosclerotic cardiovascular disease (ASCVD) and overweight/obesity but without diabetes who initiated semaglutide 2.4 mg in real‐world settings.
Materials and Methods
Individuals initiating semaglutide 2.4 mg aged ≥45 years with ASCVD and overweight/obesity but without diabetes were identified in a US database (01/01/2016–12/31/2023) and matched 1:2 to those not on semaglutide based on a non‐parsimonious propensity‐score model. The primary outcomes included revised 3‐point MACE (rMACE‐3: myocardial infarction, stroke, and all‐cause mortality) and revised 5‐point MACE (rMACE‐5: rMACE‐3, coronary revascularisation, and hospitalisation for heart failure). Secondary outcomes included MACE‐3 and MACE‐5, defined similarly to rMACE‐3 and rMACE‐5 but replacing all‐cause mortality with cardiovascular‐related mortality. Exploratory outcomes included incident type 2 diabetes, major adverse kidney events, and major obesity‐related adverse events.
Results
A total of 9321 individuals on semaglutide 2.4 mg were matched to 18,642 individuals not on semaglutide; patient characteristics were well‐balanced between cohorts. Over a mean follow‐up of 200 days, semaglutide 2.4 mg was associated with significantly lower risks of rMACE‐5 (hazard ratio: 0.55; p < 0.001), rMACE‐3 (0.43; p < 0.001), MACE‐5 (0.65; p < 0.001), and MACE‐3 (0.58; p < 0.01). Semaglutide 2.4 mg was also associated with lower risks of all‐cause mortality, cardiovascular‐related mortality, hospitalisation for heart failure, and all exploratory outcomes.
Conclusions
In this real‐world study of US individuals with ASCVD and overweight/obesity but without diabetes, semaglutide 2.4 mg was associated with significantly reduced risk of MACEs and other obesity‐related morbidities (NCT06874751).
Keywords: cardiovascular disease, cohort study, GLP‐1, observational study, real‐world evidence, semaglutide
1. INTRODUCTION
Individuals with overweight/obesity face a higher risk of developing atherosclerotic cardiovascular disease (ASCVD) compared with those with a healthy weight. 1 , 2 Consequently, an increased risk of subsequent major adverse cardiovascular events (MACE) and death is observed in this growing population. 3 The annual obesity‐related cardiovascular disease deaths in the United States (US) tripled from 2.2 per 100,000 population to 6.6 per 100,000 population between 1999 and 2020. 4
In 2024, semaglutide 2.4 mg, a glucagon‐like peptide‐1 (GLP‐1) receptor agonist indicated for weight management, was approved by the US Food and Drug Administration (FDA) for MACE risk reduction in adults with overweight/obesity and cardiovascular disease (CVD). 5 This approval was based on the results of the multicenter, double‐blind, randomised, placebo‐controlled SELECT trial, which demonstrated that once‐weekly subcutaneous semaglutide 2.4 mg was superior to placebo in reducing the incidence of death from cardiovascular causes, non‐fatal myocardial infarction (MI), or non‐fatal stroke (hazard ratio [HR], 0.80; 95% confidence interval [CI], 0.72 to 0.90; p < 0.001), in patients with ASCVD and overweight/obesity but without diabetes. 6
While there is compelling evidence from the SELECT trial supporting the efficacy of semaglutide 2.4 mg in lowering the risk of cardiovascular events among patients with ASCVD and overweight/obesity but without diabetes, real‐world data are essential for gaining insights into these effects in everyday clinical practice. Such data would provide insight into the variability in patient characteristics and management strategies that often differ from those in randomised controlled trials. This interim analysis from the observational SCORE study (Semaglutide Effects on Cardiovascular Outcomes in People with Overweight or Obesity in the Real World) aims to address this gap by using a large real‐world database to evaluate the association of semaglutide 2.4 mg use vs. no use and the risk of cardiovascular events and other obesity‐related outcomes in US adults with ASCVD and overweight/obesity but without diabetes.
2. MATERIALS AND METHODS
2.1. Study population and design
This retrospective, longitudinal, observational study used Komodo Research Data (KRD), which includes de‐identified administrative medical and pharmacy claims linked with clinical and laboratory measurements for a nationally representative US population (01/01/2016–12/31/2023). 7 , 8 , 9 Adults meeting the following criteria were included: (1) overweight or obesity, (2) diagnosed with ASCVD and (3) aged ≥45 years (Figure 1). The eligibility date was defined as the latest of first observed overweight/obesity status, first ASCVD diagnosis, or the date the patient reached 45 years of age. Overweight/obesity was identified using the International Classification of Diseases, 10th Revision, Clinical Modification (ICD‐10‐CM) diagnosis codes or recorded laboratory measurements indicating body mass index (BMI) ≥27.0 kg/m2. ASCVD was identified using ICD‐10‐CM diagnosis codes at any position in any claim for the following conditions: MI, ischemic stroke, and peripheral arterial disease (PAD). Evidence of PAD also included records of peripheral arterial revascularisation procedure identified using the ICD‐10, Procedure Coding System (ICD‐10‐PCS), Common Procedural Terminology (CPT), or Healthcare Common Procedure Coding System (HCPCS) codes. All diagnosis, procedure, and pharmacy codes used in the study were listed in Tables S1–S5.
FIGURE 1.
Patient flow chart for the overall study population of semaglutide 2.4 mg use versus non‐use. ASCVD, atherosclerotic cardiovascular disease.
Individuals who met the above criteria were sorted into two groups—those who initiated semaglutide 2.4 mg (semaglutide 2.4 mg use) and those who did not (non‐use) (Figure 1). The index date for individuals with semaglutide 2.4 mg use was defined as the initiation of semaglutide 2.4 mg identified in a pharmacy claim on or after June 4, 2021 (FDA approval date) and the eligibility date. The index date for each individual not on semaglutide was a randomly selected date on or after June 4, 2021 and the eligibility date, chosen from all the dates when the patient had at least one pharmacy claim. All patients were required to have at least 12 months of continuous insurance enrolment prior to the index date (baseline period) and to have re‐confirmation of overweight/obesity status during the baseline period.
Individuals were excluded from the study if they met any of the following criteria during the baseline period: (1) evidence of diabetes (based on ≥2 diagnoses of type 1 diabetes or ≥2 diagnoses of type 2 diabetes (T2D), a pharmacy claim of a glucose‐lowering agent indicated for diabetes, or a record of HbA1c ≥6.5%); (2) use of a GLP‐1 or GLP‐1/glucose‐dependent insulinotropic polypeptide (GIP) receptor agonist approved for weight management (excluding semaglutide 2.4 mg); (3) pregnancy in patients with a recorded female sex; (4) history of bariatric surgery; (5) diagnosis of chronic or acute pancreatitis, multiple endocrine neoplasia type 2, or medullary thyroid carcinoma at any position of a claim record; (6) end‐stage kidney disease (ESKD), defined based on ICD‐10‐CM diagnosis codes, an estimated glomerular filtration rate (eGFR) <15 mL/min/1.73m2, a procedure for ESKD, chronic or intermittent haemodialysis or peritoneal dialysis, or kidney transplant.
Eligible individuals were followed from the index date to the end of their continuous enrolment period, end of data availability, initiation of a non‐semaglutide 2.4 mg GLP‐1 or GLP‐1/GIP receptor agonist, a bariatric surgical procedure, or death, whichever occurred first.
2.2. Study variables and outcomes
Patients' demographic and clinical characteristics, comorbidities, relevant procedures received, medication use, and healthcare resource utilisation were evaluated during the baseline period and/or on the index date (Table 1).
TABLE 1.
Patient baseline characteristics of semaglutide 2.4 mg use versus non‐use before and after 1:2 propensity score matching.
| Pre‐PS matching | Post‐PS matching (up to 1:2) a | |||||
|---|---|---|---|---|---|---|
| Semaglutide 2.4 mg use | Non‐use | SMD | Semaglutide 2.4 mg use | Non‐use | SMD | |
| N = 9322 | N = 1 328 554 | N = 9321 | N = 18 642 | |||
| Demographics | ||||||
| Sex, n (%) | ||||||
| Male | 2846 (30.5%) | 633 009 (47.6%) | 0.36 | 2846 (30.5%) | 5804 (31.1%) | 0.01 |
| Female | 6367 (68.3%) | 672 833 (50.6%) | 0.37 | 6366 (68.3%) | 12 586 (67.5%) | 0.02 |
| Undefined | 109 (1.2%) | 22 712 (1.7%) | 0.05 | 109 (1.2%) | 252 (1.4%) | 0.02 |
| Age at index in years, mean ± SD [median] | 56.8 ± 7.1 [56.0] | 66.3 ± 11.4 [65.0] | 0.99 | 56.8 ± 7.1 [56.0] | 57.1 ± 7.3 [57.0] | 0.03 |
| Age at index in years, n (%) | ||||||
| ≥45 and ≤55 | 4243 (45.5%) | 258 428 (19.5%) | 0.58 | 4242 (45.5%) | 8529 (45.8%) | 0.00 |
| ≥56 and ≤65 | 4174 (44.8%) | 414 474 (31.2%) | 0.28 | 4174 (44.8%) | 8402 (45.1%) | 0.01 |
| ≥66 and ≤75 | 772 (8.3%) | 337 514 (25.4%) | 0.47 | 772 (8.3%) | 1517 (8.1%) | 0.01 |
| ≥76 and ≤85 | 130 (1.4%) | 236 721 (17.8%) | 0.58 | 130 (1.4%) | 150 (0.8%) | 0.06 |
| ≥86 | 3 (0.0%) | 81 417 (6.1%) | 0.36 | 3 (0.0%) | 44 (0.2%) | 0.06 |
| Race/ethnicity, n (%) | ||||||
| White | 4603 (49.4%) | 729 143 (54.9%) | 0.11 | 4602 (49.4%) | 9068 (48.6%) | 0.01 |
| Black or African American | 1196 (12.8%) | 188 005 (14.2%) | 0.04 | 1196 (12.8%) | 2367 (12.7%) | 0.00 |
| Hispanic or Latino | 980 (10.5%) | 190 944 (14.4%) | 0.12 | 980 (10.5%) | 1979 (10.6%) | 0.00 |
| Other | 406 (4.4%) | 42 726 (3.2%) | 0.06 | 406 (4.4%) | 831 (4.5%) | 0.01 |
| Asian or Pacific Islander | 131 (1.4%) | 30 518 (2.3%) | 0.07 | 131 (1.4%) | 268 (1.4%) | 0.00 |
| Unknown | 2006 (21.5%) | 147 218 (11.1%) | 0.29 | 2006 (21.5%) | 4129 (22.1%) | 0.02 |
| Geographic region, n (%) | ||||||
| Midwest | 1993 (21.4%) | 274 334 (20.6%) | 0.02 | 1993 (21.4%) | 3886 (20.8%) | 0.01 |
| South | 3168 (34.0%) | 437 590 (32.9%) | 0.02 | 3168 (34.0%) | 6418 (34.4%) | 0.01 |
| Northeast | 2807 (30.1%) | 397 128 (29.9%) | 0.00 | 2806 (30.1%) | 5529 (29.7%) | 0.01 |
| West | 1350 (14.5%) | 212 259 (16.0%) | 0.04 | 1350 (14.5%) | 2785 (14.9%) | 0.01 |
| Unknown | 4 (0.0%) | 7243 (0.5%) | 0.09 | 4 (0.0%) | 24 (0.1%) | 0.03 |
| Insurance coverage b , n (%) | ||||||
| Commercial | 7248 (77.8%) | 454 220 (34.2%) | 0.98 | 7247 (77.7%) | 14 611 (78.4%) | 0.02 |
| Medicaid | 1355 (14.5%) | 283 810 (21.4%) | 0.18 | 1355 (14.5%) | 2710 (14.5%) | 0.00 |
| Medicare | 710 (7.6%) | 664 801 (50.0%) | 1.06 | 710 (7.6%) | 1254 (6.7%) | 0.03 |
| Unknown | 250 (2.7%) | 28 085 (2.1%) | 0.04 | 250 (2.7%) | 496 (2.7%) | 0.00 |
| Study design characteristics | ||||||
| Index year, n (%) | ||||||
| 2021 | 649 (7.0%) | 278 729 (21.0%) | 0.41 | 649 (7.0%) | 1160 (6.2%) | 0.03 |
| 2022 | 998 (10.7%) | 506 394 (38.1%) | 0.67 | 998 (10.7%) | 2070 (11.1%) | 0.01 |
| 2023 | 7675 (82.3%) | 543 431 (40.9%) | 0.94 | 7674 (82.3%) | 15 412 (82.7%) | 0.01 |
| Duration between eligibility date and index date (months), mean ± SD [median] | 34.1 ± 24.8 [29.5] | 29.2 ± 24.6 [23.0] | 0.20 | 34.1 ± 24.8 [29.5] | 33.8 ± 26.8 [28.5] | 0.01 |
| Cardiometabolic markers closest to index date | ||||||
| Obesity class, n (%) c | ||||||
| Patients with BMI category information | 3289 (35.3%) | 279 805 (21.1%) | 0.32 | 3289 (35.3%) | 6539 (35.1%) | 0.00 |
| Overweight | 246 (2.6%) | 55 553 (4.2%) | 0.09 | 246 (2.6%) | 493 (2.6%) | 0.00 |
| Obese | 3043 (32.6%) | 224 252 (16.9%) | 0.37 | 3043 (32.6%) | 6046 (32.4%) | 0.00 |
| Patients with detailed BMI value information | 6033 (64.7%) | 1 048 749 (78.9%) | 0.32 | 6032 (64.7%) | 12 103 (64.9%) | 0.00 |
| BMI, mean ± SD [median] | 37.3 ± 5.1 [37.3] | 32.8 ± 4.8 [31.5] | 0.90 | 37.3 ± 5.1 [37.3] | 37.1 ± 5.1 [37.0] | 0.03 |
| Overweight (≥27 to <30) | 382 (4.1%) | 353 049 (26.6%) | 0.66 | 382 (4.1%) | 673 (3.6%) | 0.03 |
| Obesity class 1 (≥30 to <35) | 1798 (19.3%) | 399 989 (30.1%) | 0.25 | 1798 (19.3%) | 3595 (19.3%) | 0.00 |
| Obesity class 2 (≥35 to <40) | 1698 (18.2%) | 173 517 (13.1%) | 0.14 | 1698 (18.2%) | 3522 (18.9%) | 0.02 |
| Obesity class 3 (≥40) | 2155 (23.1%) | 122 194 (9.2%) | 0.39 | 2154 (23.1%) | 4313 (23.1%) | 0.00 |
| Evidence of smoking history, n (%) | 2110 (22.6%) | 415 602 (31.3%) | 0.20 | 2110 (22.6%) | 4261 (22.9%) | 0.01 |
| ASCVD before index date, n (%) | ||||||
| MI | 3177 (34.1%) | 457 771 (34.5%) | 0.01 | 3177 (34.1%) | 6382 (34.2%) | 0.00 |
| Ischemic stroke | 2882 (30.9%) | 408 206 (30.7%) | 0.00 | 2882 (30.9%) | 5619 (30.1%) | 0.02 |
| PAD | 4157 (44.6%) | 694 262 (52.3%) | 0.15 | 4156 (44.6%) | 8234 (44.2%) | 0.01 |
| Comorbidities during baseline period, n (%) | ||||||
| CV and renal | ||||||
| Hypertension | 6619 (71.0%) | 1 037 299 (78.1%) | 0.16 | 6618 (71.0%) | 13 337 (71.5%) | 0.01 |
| Dyslipidaemia | 6338 (68.0%) | 986 040 (74.2%) | 0.14 | 6337 (68.0%) | 12 679 (68.0%) | 0.00 |
| HF | 743 (8.0%) | 214 745 (16.2%) | 0.25 | 743 (8.0%) | 1529 (8.2%) | 0.01 |
| Unstable angina | 649 (7.0%) | 142 188 (10.7%) | 0.13 | 649 (7.0%) | 1320 (7.1%) | 0.00 |
| A‐Fib | 733 (7.9%) | 201 990 (15.2%) | 0.23 | 733 (7.9%) | 1426 (7.6%) | 0.01 |
| CKD d | 463 (5.0%) | 202 114 (15.2%) | 0.35 | 463 (5.0%) | 952 (5.1%) | 0.01 |
| Other ORC | ||||||
| Musculoskeletal pain e | 4573 (49.1%) | 559 435 (42.1%) | 0.14 | 4572 (49.1%) | 9060 (48.6%) | 0.01 |
| Prediabetes | 3457 (37.1%) | 395 309 (29.8%) | 0.16 | 3456 (37.1%) | 6823 (36.6%) | 0.01 |
| OSA | 3223 (34.6%) | 276 408 (20.8%) | 0.31 | 3222 (34.6%) | 6397 (34.3%) | 0.01 |
| GERD | 3049 (32.7%) | 430 298 (32.4%) | 0.01 | 3048 (32.7%) | 5980 (32.1%) | 0.01 |
| Knee osteoarthritis | 1678 (18.0%) | 207 498 (15.6%) | 0.06 | 1677 (18.0%) | 3314 (17.8%) | 0.01 |
| Asthma | 1578 (16.9%) | 166 287 (12.5%) | 0.12 | 1578 (16.9%) | 3100 (16.6%) | 0.01 |
| MASH/MAFLD | 902 (9.7%) | 79 328 (6.0%) | 0.14 | 902 (9.7%) | 1770 (9.5%) | 0.01 |
| MASH | 90 (1.0%) | 6422 (0.5%) | 0.06 | 90 (1.0%) | 190 (1.0%) | 0.01 |
| MAFLD | 849 (9.1%) | 75 809 (5.7%) | 0.13 | 849 (9.1%) | 1688 (9.1%) | 0.00 |
| Procedures during baseline period, n (%) | ||||||
| Coronary revascularisation | 172 (1.8%) | 43 258 (3.3%) | 0.09 | 172 (1.8%) | 360 (1.9%) | 0.01 |
| Peripheral arterial revascularisation | 26 (0.3%) | 10 754 (0.8%) | 0.07 | 26 (0.3%) | 75 (0.4%) | 0.02 |
| Carotid intervention | 9 (0.1%) | 4172 (0.3%) | 0.05 | 9 (0.1%) | 15 (0.1%) | 0.01 |
| Medication usage during baseline period, n (%) | ||||||
| Statin therapy | 4800 (51.5%) | 816 676 (61.5%) | 0.20 | 4799 (51.5%) | 9656 (51.8%) | 0.01 |
| Beta‐blocker | 3521 (37.8%) | 588 548 (44.3%) | 0.13 | 3521 (37.8%) | 7144 (38.3%) | 0.01 |
| ARBs | 2681 (28.8%) | 369 065 (27.8%) | 0.02 | 2680 (28.8%) | 5405 (29.0%) | 0.01 |
| Thiazides | 2592 (27.8%) | 329 698 (24.8%) | 0.07 | 2591 (27.8%) | 5161 (27.7%) | 0.00 |
| Calcium channel blockers | 2378 (25.5%) | 419 879 (31.6%) | 0.14 | 2378 (25.5%) | 4726 (25.4%) | 0.00 |
| ACE inhibitors | 1979 (21.2%) | 349 088 (26.3%) | 0.12 | 1979 (21.2%) | 4007 (21.5%) | 0.01 |
| Loop diuretics | 1259 (13.5%) | 222 109 (16.7%) | 0.09 | 1259 (13.5%) | 2555 (13.7%) | 0.01 |
| Anticoagulants f | 1033 (11.1%) | 218 609 (16.5%) | 0.16 | 1033 (11.1%) | 2075 (11.1%) | 0.00 |
| Antiplatelet agents and dual antiplatelet agents | 999 (10.7%) | 200 503 (15.1%) | 0.13 | 999 (10.7%) | 2026 (10.9%) | 0.00 |
| Anti‐obesity medications g | 938 (10.1%) | 13 939 (1.0%) | 0.40 | 937 (10.1%) | 1610 (8.6%) | 0.05 |
| Aldosterone antagonists | 33 (0.4%) | 2689 (0.2%) | 0.03 | 33 (0.4%) | 52 (0.3%) | 0.01 |
| HRU during baseline period | ||||||
| Number of IP visits, mean ± SD [median] | 0.14 ± 0.45 [0.00] | 0.29 ± 0.73 [0.00] | 0.26 | 0.14 ± 0.45 [0.00] | 0.14 ± 0.43 [0.00] | 0.01 |
| Number of ER visits, mean ± SD [median] | 1.17 ± 3.25 [0.00] | 1.42 ± 5.94 [0.00] | 0.05 | 1.17 ± 3.25 [0.00] | 1.18 ± 2.94 [0.00] | 0.00 |
| Number of OP visits, mean ± SD [median] | 26.5 ± 30.20 [19.00] | 28.7 ± 42.17 [18.00] | 0.06 | 26.5 ± 30.20 [19.00] | 26.0 ± 37.27 [17.00] | 0.02 |
| Receipt of weight management consultation and specialty care service, n (%) | 1658 (17.8%) | 128 053 (9.6%) | 0.24 | 1657 (17.8%) | 3291 (17.7%) | 0.00 |
Abbreviations: ACE, angiotensin‐converting enzyme; A‐Fib, atrial fibrillation; ARB, angiotensin II receptor blocker; ASCVD, atherosclerotic cardiovascular disease; BMI, body mass index; CKD, chronic kidney disease; CV, cardiovascular; ER, emergency room; GERD, gastroesophageal reflux disease; HF, heart failure; HRU, healthcare resource utilisation; IP, inpatient; MAFLD, metabolic dysfunction‐associated fatty liver disease; MASH, metabolic dysfunction‐associated steatohepatitis; MI, myocardial infarction; ORC, obesity‐related comorbidity; OP, outpatient; OSA, obstructive sleep apnoea; PAD, peripheral arterial disease; PS, propensity score; SD, standard deviation; SMD, standardised mean difference.
Based on the propensity scores, each individual on semaglutide 2.4 mg was matched to two individuals who were not on semaglutide using greedy nearest neighbour propensity score matching within a calliper without replacement. The calliper width threshold was 0.2 of the pooled standard deviation of the logit of the propensity score of individuals on semaglutide 2.4 mg and those who were not on it.
Includes primary and available secondary insurance plans on the index date.
‘Patients with BMI category information’ refers to individuals without exact BMI values available; ‘patients with detailed BMI value information’ refers to individuals with exact BMI values available.
Identified as Stage 3 or 4 CKD or unspecified stage CKD.
Includes pain of knee, hip, and spine.
Including warfarin, thrombin inhibitors, and direct factor Xa inhibitors.
Including orlistat, phentermine‐topiramate, phentermine, and naltrexone‐bupropion.
The primary outcomes included the revised 5‐point and 3‐point major cardiovascular adverse events (rMACE‐5 and rMACE‐3, respectively). rMACE‐5 was defined as any occurrence of MI, stroke (ischemic or haemorrhagic), hospitalisation for heart failure (HF), coronary revascularisation, or all‐cause mortality. rMACE‐3 was defined as any occurrence of MI, stroke, or all‐cause mortality. 10 , 11 MI, stroke, and hospitalisation for HF were identified for the corresponding conditions with ICD‐10‐CM diagnosis code in the primary position during an inpatient visit. Coronary revascularisation was identified using ICD‐10‐PCS, CPT, or HCPCS procedure codes in any claim. Death was recorded at the month and year level in KRD, with the last day of the recorded month of death used as the event date.
Secondary outcomes included MACE‐5 and MACE‐3, defined similarly to rMACE‐3 and rMACE‐5 but replacing all‐cause mortality with cardiovascular‐related mortality. Cardiovascular‐related mortality was defined as a death event with any claim within 30 days before death 12 , 13 with a primary diagnosis of MI, ischemic or haemorrhagic stroke, HF, cardiovascular haemorrhage, other cardiovascular conditions, or evidence of peripheral arterial revascularisation, coronary revascularisation, or carotid intervention. rMACE‐5 and rMACE‐3 with all‐cause mortality were included as primary outcomes, while MACE‐5 and MACE‐3 with cardiovascular‐related mortality were included as secondary outcomes, due to potential challenges in identifying cardiovascular‐related mortality using real‐world claims data.
Exploratory outcomes included (1) incident T2D, defined as ≥2 diagnoses of T2D on distinct dates based on ICD‐10‐CM diagnosis codes or a record of HbA1c ≥6.5%; (2) major adverse kidney events (MAKE), defined as any occurrence of ESKD, acute kidney injury (AKI), or all‐cause mortality; and (3) major obesity‐related adverse events (MORAE), defined as any occurrence of rMACE‐5, incident T2D, MAKE, or all‐cause hospitalisation. Individual components of all composite outcomes were also analysed.
2.3. Statistical analysis
A non‐parsimonious propensity‐score model was developed to alleviate the effect of potential confounding. Variables that may have affected cohort assignment or outcomes, that is, all variables evaluated during the baseline period and/or on the index date (Tables 1 and S6), were included in the model. Based on the propensity scores, each individual on semaglutide 2.4 mg was matched to two individuals who were not on semaglutide. Characteristics of the cohort were described by semaglutide 2.4 mg use and non‐use, both before and after matching, using standardised mean difference (SMD) to evaluate covariate balance. An SMD <0.1 indicates well‐balanced characteristics between cohorts. 14
Cumulative incidence functions (CIFs) estimated by Aalen‐Johansen estimators, which account for competing risks, were visualised for both individuals on semaglutide 2.4 mg and those not on it for the first event of the study outcomes. 15 , 16 For the outcomes of MACE‐5, MACE‐3, and cardiovascular‐related mortality, non‐cardiovascular‐related mortality (i.e., any death not classified as cardiovascular‐related) was considered as a competing risk. For individual component outcomes other than all‐cause mortality and cardiovascular‐related mortality, all‐cause mortality was considered as a competing risk. Cox proportional hazards (PH) models were used to compare time to the first event for the study outcomes between individuals on semaglutide 2.4 mg and those who were not on it. HRs, 95% CIs, and p‐values were reported.
Several additional sensitivity analyses were conducted: first, we calculated the E‐values, which assess how strong an unmeasured confounder would have to be to explain away an observed association between semaglutide use and the outcomes. Second, the analyses for primary and secondary outcomes were repeated using an ‘as‐treated’ analysis, where individuals on semaglutide 2.4 mg were censored at the date of treatment discontinuation, defined as the last day of supply without a subsequent prescription within 30 days. Individuals not on semaglutide were censored at the date corresponding to the same follow‐up duration as their matched counterparts. Third, we assessed the association between semaglutide 2.4 mg use and sensorineural hearing loss as a negative control outcome. Fourth, the analyses of the primary and secondary outcomes were repeated in two different patient populations: (1) patients with detailed BMI value information or (2) patients who had no primary inpatient diagnosis of MI, stroke, unstable angina, or transient ischaemic attack in 60 days prior to the index date. Finally, we conducted subgroup analyses stratified by age (≤55 years vs. >55 years), sex (male vs. female), and race (patients of white race vs. patients of non‐white race).
A two‐tailed p < 0.05 was considered statistically significant for all analyses. No corrections for multiple comparisons have been performed due to the potentially high correlation among study outcomes. 17
3. RESULTS
3.1. Characteristics of the study population
A total of 9322 individuals who were on semaglutide 2.4 mg and 1 328 554 individuals who were not on it were identified (Figure 1). Before matching, individuals on semaglutide 2.4 mg were younger than those not on it (mean [SD] 56.8 [7.1] vs. 66.3 [11.4] years), more frequently female (68.3% vs. 50.6%), had a higher prevalence of obesity (93.3% vs. 69.2%) and were more likely to have commercial insurance coverage (77.8% vs. 34.2%) (Table 1). Individuals on semaglutide 2.4 mg were less likely to have HF (8.0% vs. 16.2%), chronic kidney disease (5.0% vs. 15.2%), and had lower use of cardiovascular disease therapies but higher use of obesity medications (10.1% vs. 1.0%) during the baseline period.
All but one individual on semaglutide 2.4 mg were matched to two individuals not on semaglutide, resulting in 18,642 matches that were not taking the drug. The patient characteristics of the two cohorts were well‐balanced after propensity score matching, with SMD ≤0.1 for all characteristics (Table 1). After matching, the mean age of the study population was 57.0 (SD: 7.2) years, 67.8% were female, and most had commercial insurance (78.2%). The racial/ethnic distribution included 48.9% identified as White, 12.7% as Black/African American, and 10.6% as Hispanic/Latino. Race/ethnicity was classified as unknown for 21.9% of participants. The majority of patients (93.6%) were classified as having obesity, and among patients with detailed BMI value information (64.9%), the mean BMI was 37.2 (SD: 5.1) kg/m2. Prior to the index date, ASCVD identifiers of MI, ischemic stroke, and PAD were observed among 34.2%, 30.4%, and 44.3% of patients, respectively.
During the 12‐month baseline period, the most commonly observed cardiovascular and renal comorbidities were hypertension (71.4%) and dyslipidaemia (68.0%). The most commonly recorded other obesity‐related comorbidities during baseline were musculoskeletal pain (48.8%), prediabetes (36.8%), obstructive sleep apnoea (34.4%), and gastroesophageal reflux disease (32.3%). The use of at least one medical therapy for CVD was frequently observed during baseline.
3.2. Primary outcomes (rMACE‐3 and rMACE‐5)
In the matched individuals who were on semaglutide 2.4 mg vs. those who were not, there were 88 and 288 rMACE‐5 events and 42 and 175 rMACE‐3 events, respectively (Table 2). The corresponding incidence rates per 1000 person‐years were 16.0 vs. 29.4 for rMACE‐5 and 7.6 vs. 17.8 for rMACE‐3, respectively. Compared with non‐use, semaglutide 2.4 mg use was associated with a significantly lower risk of rMACE‐5 by 45% (HR [95% CI]: 0.55 [0.43–0.69]; p < 0.001) and of rMACE‐3 by 57% (HR: 0.43 [0.31–0.61]; p < 0.001), over a mean follow‐up of 216 days (semaglutide 2.4 mg use) and 195 days (non‐use) (Figure 2 and Table 2). In addition, compared with non‐use, semaglutide 2.4 mg use was associated with a significantly lower risk of hospitalisation for HF and all‐cause mortality (Table 2). The cumulative incidence of individual components of the composite outcomes is displayed in Figure S1.
TABLE 2.
Number of events, incidence rates, and HR for study outcomes in matched semaglutide 2.4 mg use versus non‐use.
| Outcomes | Descriptive analysis | Comparative analysis (semaglutide 2.4 mg use vs. non‐use) | ||||
|---|---|---|---|---|---|---|
| Semaglutide 2.4 mg use | Non‐use | |||||
| N = 9321 | N = 18 642 | |||||
| n (%) | IR per 1000 patient‐years | n (%) | IR per 1000 patient‐years | HR a (95% CI) | p‐value | |
| Primary outcomes | ||||||
| rMACE‐5 b | 88 (0.94%) | 16.00 | 288 (1.55%) | 29.38 | 0.55 (0.43, 0.69) | <0.001* |
| MI | 19 (0.20%) | 3.44 | 50 (0.27%) | 5.06 | 0.69 (0.41, 1.18) | 0.175 |
| Stroke c | 16 (0.17%) | 2.90 | 45 (0.24%) | 4.56 | 0.65 (0.37, 1.16) | 0.144 |
| Hospitalisation for HF | 22 (0.24%) | 3.98 | 87 (0.47%) | 8.83 | 0.46 (0.29, 0.73) | <0.01* |
| Coronary revascularisation d | 39 (0.42%) | 7.07 | 75 (0.40%) | 7.61 | 0.93 (0.63, 1.36) | 0.696 |
| All‐cause mortality | 7 (0.08%) | 1.27 | 90 (0.48%) | 9.09 | 0.14 e (0.06, 0.30) | <0.001* |
| rMACE‐3 f | 42 (0.45%) | 7.61 | 175 (0.94%) | 17.75 | 0.43 (0.31, 0.61) | <0.001* |
| Secondary outcomes | ||||||
| MACE‐5 g | 87 (0.93%) | 15.82 | 242 (1.30%) | 24.69 | 0.65 (0.51, 0.83) | <0.001* |
| MACE‐3 h | 41 (0.44%) | 7.43 | 128 (0.69%) | 12.98 | 0.58 (0.41, 0.83) | <0.01* |
| CV‐related mortality | 6 (0.06%) | 1.08 | 40 (0.22%) | 4.04 | 0.27 (0.11, 0.63) | <0.01* |
| Exploratory outcomes | ||||||
| Incident T2D | 200 (2.15%) | 36.81 | 451 (2.42%) | 46.85 | 0.80 (0.68, 0.94) | <0.01* |
| MAKE | 29 (0.31%) | 5.25 | 171 (0.92%) | 17.35 | 0.30 (0.21, 0.45) | <0.001* |
| ESKD i | 8 (0.09%) | 1.45 | 26 (0.14%) | 2.63 | 0.55 (0.25, 1.21) | 0.139 |
| All‐cause mortality | 7 (0.08%) | 1.27 | 90 (0.48%) | 9.09 | 0.14 (0.06, 0.30) | <0.001* |
| AKI j | 15 (0.16%) | 2.71 | 72 (0.39%) | 7.30 | 0.38 (0.22, 0.66) | <0.001* |
| MORAE k | 641 (6.88%) | 122.09 | 1357 (7.28%) | 146.64 | 0.84 (0.77, 0.92) | <0.001* |
| All‐cause hospitalisation | 442 (4.74%) | 82.62 | 915 (4.91%) | 96.35 | 0.87 (0.77, 0.97) | <0.05* |
Abbreviations: AKI, acute kidney injury; CABG, coronary artery bypass grafting; CI, confidence interval; CV, cardiovascular; eGFR, estimated glomerular filtration rate; ESKD, end‐stage kidney disease; HF, heart failure; HR, hazard ratio; IR, incidence rate; MACE‐3, 3‐point major adverse cardiovascular events; MACE‐5, 5‐point major adverse cardiovascular events; MAKE, major adverse kidney events; MI, myocardial infarction; MORAE, major obesity‐related adverse events; PCI, percutaneous coronary intervention; rMACE‐3, revised MACE‐3; rMACE‐5, revised MACE‐5; T2D, type 2 diabetes. * Indicates statistical signifinance at p 〈 0.05.
Cox proportional hazards regression models were used to compare study outcomes between semaglutide 2.4 mg use versus non‐use.
Composite of the occurrence of MI, stroke, hospitalisation for HF, coronary revascularisation, and all‐cause mortality.
Includes both ischemic and haemorrhagic stroke.
Evidence of coronary revascularisation procedure, defined as either a PCI or CABG, in any claim.
The PH assumption was not met for all‐cause mortality (Schoenfeld residual test: p < 0.05).
Composite of MI, stroke, and all‐cause mortality.
Composite of MI, stroke, hospitalisation for HF, coronary revascularisation, and CV‐related mortality.
Composite of MI, stroke, and CV‐related mortality.
Composite of ESKD‐related diagnosis or services, a record of eGFR <15 mL/min/1.73m2, incident chronic dialysis, and incident kidney transplant.
AKI was defined using ICD‐10‐CM diagnosis code in the primary position during an inpatient visit.
Composite of MACE‐5, incident T2D, MAKE, and all‐cause hospitalisation.
FIGURE 2.
Cumulative incidence of primary outcomes among matched semaglutide 2.4 mg use versus non‐use for time to (A) rMACE‐5 and (B) rMACE‐3. CI, confidence interval; HR, hazard ratio; rMACE‐3, revised 3‐point major adverse cardiovascular events; rMACE‐5, revised 5‐point major adverse cardiovascular events.
3.3. Secondary outcomes (MACE‐5 and MACE‐3)
For MACE‐5 and MACE‐3, the numbers of events for the matched individuals on semaglutide 2.4 mg vs. those who were not on it were 87 (incidence rate per 1000 person‐years: 15.8) vs. 242 (24.7) and 41 (7.4) vs. 128 (13.0), respectively (Table 2; Figure 3). Compared with non‐use, semaglutide 2.4 mg use was associated with a significantly lower risk of MACE‐5 by 35% (HR: 0.65 [0.51–0.83]; p < 0.001) and of MACE‐3 by 42% (HR: 0.58 [0.41–0.83]; p < 0.01). A consistent effect was also observed for cardiovascular‐related mortality (Table 2).
FIGURE 3.
Cumulative incidence of secondary outcomes among matched semaglutide 2.4 mg use versus non‐use for time to (A) MACE‐5 and (B) MACE‐3. CI, confidence interval; HR, hazard ratio; MACE‐3, 3‐point major adverse cardiovascular events; MACE‐5, 5‐point major adverse cardiovascular events.
3.4. Exploratory outcomes
Compared with non‐use, semaglutide 2.4 mg use was associated with a significantly lower risk of incident T2D (HR: 0.80 [0.68–0.94]; p < 0.01; Table 2; Figure S2), MAKE (HR: 0.30 [0.21–0.45]; p < 0.001; Table 2; Figure S2), AKI (HR: 0.38 [0.22–0.66]; p < 0.001; Table 2; Figure S3), MORAE (HR: 0.84 [0.77–0.92]; p < 0.001; Table 2; Figure S2), and all‐cause hospitalisation (HR: 0.87 [0.77–0.97]; p < 0.05; Table 2; Figure S3).
3.5. Sensitivity analyses
Similar results were observed using the ‘as‐treated’ approach (Table S7), in patients with detailed BMI value information (Table S8) and in patients with no MI, stroke, unstable angina, or transient ischemic attack in 60 days prior to the index date (Table S9). Subgroup analysis by age, sex, and race also produced consistent results (Figure S4). The negative control outcome showed no difference between study cohorts (HR: 1.03 [0.87–1.21]; p = 0.775; Table S10). Large E‐values for rMACE‐5 and rMACE‐3 were 3.0 and 4.1, respectively, suggesting robustness of the treatment–outcome association.
4. DISCUSSION
In this large retrospective study of 27,963 US individuals with ASCVD and overweight/obesity but without diabetes from real‐world clinical practice, semaglutide 2.4 mg use was associated with significantly lower risks of all MACE outcomes, supporting the significant clinical cardiovascular benefits observed in the SELECT trial. Notably lower risks were also observed across individual components of the MACE outcomes. Additionally, the results of the exploratory analyses were consistent with the data from clinical trials for semaglutide 2.4 mg in terms of lower risk of obesity‐related outcomes such as incident T2D and kidney outcomes, underscoring its broader therapeutic effects.
The substantial and early reduction in CV risks associated with semaglutide 2.4 mg observed in this study is consistent with the findings from the SELECT trial at similar time points. A recent secondary analysis of the SELECT trial demonstrated that the nominal significant benefit of semaglutide 2.4 mg for MACE was observed as early as 3 months post‐randomisation and over the first 6 months. 18 The association between semaglutide 2.4 mg in real‐world practice and cardiovascular benefits demonstrated in this study is remarkably similar to those seen in the SELECT trial (MACE‐3 HR: 0.59 [0.44–0.80] at 6 months). This is also supported by previous studies which found that the cardiovascular benefits of semaglutide 2.4 mg appeared long before maximal weight loss was achieved. 6 , 19 This could be due to alternative vascular, inflammatory, insulin resistance, metabolic, or other mechanisms beyond weight loss that contribute to the cardioprotective effect of semaglutide 2.4 mg. A mediation analysis suggested that about 80% of the MACE risk reduction with semaglutide is mediated by factors other than body weight loss. 20 Taken together, current results combined with existing evidence suggest that the cardiovascular benefits of semaglutide 2.4 mg in patients with ASCVD and overweight/obesity but without diabetes may be driven by accelerated physiological changes beyond just the extent of weight loss. Additionally, these cardiovascular benefits of semaglutide 2.4 mg appear to be molecule‐specific and not a class effect of all GLP‐1 receptor agonists. 21 , 22
Findings from this study extend the evidence base for semaglutide 2.4 mg by demonstrating not only a reduced risk of MACE outcomes, but also a significantly lower risk of incident T2D, MAKE, and MORAE in a real‐world population. This is particularly relevant in the context of the American Heart Association's emerging cardiovascular‐kidney‐metabolic (CKM) framework, which emphasises a more integrated approach to cardiometabolic risk reduction. 23 Our results align with this paradigm shift by underscoring the potential of cardiovascular therapies such as semaglutide 2.4 mg to deliver broad clinical benefits across interrelated metabolic, renal, and cardiovascular domains, thus shifting the focus from MACE outcomes alone to a more holistic view of relevant health outcomes.
This study benefits from several strengths as the first large‐scale real‐world study investigating the cardiovascular benefits of semaglutide 2.4 mg in a diverse patient population using a nationally representative database. First, the present results confirmed the benefits of semaglutide 2.4 mg in reducing risks of cardiovascular outcomes among patients with ASCVD and overweight/obesity but without diabetes in routine clinical practice. Second, compared to the population in the SELECT trial, this study had a diverse population more representative of real‐world individuals taking semaglutide 2.4 mg, with close to two‐thirds of female patients and more than half of non‐white patients, complementing results from the SELECT trial. Consistent subgroup analysis results by age, sex, and race in this study strengthened the generalisability of findings from SELECT. Third, patients were rigorously matched using propensity score methods, similar to prior studies 10 , 24 with the goal of investigating if benefits observed in clinical trials were also observed in the real‐world setting. Patients were matched using a comprehensive list of patient characteristics based on literature review and clinical input, which allowed for maximisation of balance and minimisation of confounding between semaglutide 2.4 mg use and non‐use to ensure robust comparisons. Fourth, this study evaluated a broader set of outcomes beyond those assessed in the SELECT trial, suggesting favourable effects associated with semaglutide 2.4 on a wide range of complications, including incident T2D and MAKE, and in improving overall metabolic health. Lastly, the results were unchanged after multiple sensitivity analyses. Specifically, consistent results from the ‘as treated’ sensitivity analysis are important because adherence to semaglutide 2.4 mg in the real‐world might differ from clinical trial regimens, and discontinuation rates due to side effects or other factors could impact outcome estimates. Thus, the present study also confirms the cardiovascular benefits of semaglutide 2.4 mg in patients with adherence and persistence in routine clinical care compared to that of a clinical trial.
This study is also subject to several limitations. First, given the observational nature of the current study, despite robust propensity score matching and multiple sensitivity analyses, a possibility of residual unmeasured confounding cannot be ruled out. However, large E‐values for the primary outcomes partially mitigate this concern. For example, the observed HR of 0.55 between semaglutide use and rMACE‐5 could only be explained away by an unmeasured confounder that was simultaneously associated with semaglutide use by a risk ratio of 3 and with rMACE‐5 by a hazard ratio of 3, after conditioning on all patient characteristics included in the current propensity‐score model. In addition, semaglutide 2.4 mg use was not associated with the negative control outcome, sensorineural hearing loss, suggesting that the impact of unmeasured confounding may be limited. Second, the approval of semaglutide 2.4 mg in 2021 results in a relatively short duration of follow‐up, making it infeasible to assess outcomes that require a longer time to develop, such as cancer and neurodegenerative disease. Nevertheless, our results and the effect sizes observed will need replication and validation in other real‐world populations to expand the evidence base. The short follow‐up also limited the assessment of its possible long‐term effects. However, similar benefits at around 6 months for the reduction of MACE outcomes were also observed in the SELECT trial, suggesting that follow‐up duration in this study, which is comparable to other real‐world studies evaluating cardiovascular therapies, 10 , 25 is unlikely to affect the evaluation of cardiovascular outcomes. 6 Third, the restriction to patients with at least 12 months of continuous insurance coverage before the index date may exclude individuals with intermittent insurance coverage or those from underserved populations, limiting generalisability. Finally, this study relied on administrative claims data, which may lack specific clinical details including precise timing of comorbidities and death, clinical severity, as well as case identification, which can lead to potential misclassification of exposure, outcomes, or covariates. However, the KRD database used in this study was linked with clinical and laboratory measurements that are captured in routine clinical practice.
In conclusion, in this real‐world study of US patients with ASCVD and overweight/obesity but without diabetes, semaglutide 2.4 mg was associated with significantly lower risks of MACE and other obesity‐related complications. Further research using data with longer follow‐up is needed to extend the present findings to assess the long‐term benefits of semaglutide 2.4 mg in this patient population. Additionally, future studies of semaglutide 2.4 mg in broader patient populations, such as primary prevention among patients without ASCVD or secondary prevention among patients with a broader set of pre‐existing cardiovascular risk factors (e.g., carotid artery disease and aortic aneurysm), would provide valuable insight into the therapeutic potential of semaglutide 2.4 mg among patients with overweight/obesity.
AUTHOR CONTRIBUTIONS
Design: All authors contributed to the design of the study. Conduct/data collection: Zhenxiang Zhao, Wojciech Michalak, Mads Faurby, B. Gabriel Smolarz, Jinlin Song, Yan Chen and Joanna Boland contributed to the data collection of the study. Analysis: All authors contributed to the data analysis of the study. Writing manuscript: All authors contributed to the drafting of the manuscript.
CONFLICT OF INTEREST STATEMENT
Zhenxiang Zhao, Wojciech Michalak, Mads Faurby, and B. Gabriel Smolarz are employees of Novo Nordisk, Inc. and hold stock/options. Jinlin Song, Yan Chen, and Joanna Boland are employees of Analysis Group, Inc., which has received consulting fees from Novo Nordisk, Inc. for this work. Kim G. Smolderen receives research grants from Johnson & Johnson, Merck, and Abbott. She is a consultant for Dario Health, Terumo, Novo Nordisk, and Merck. She is the owner of BoboDream LLC. She receives funding from the National Institutes of Health (R01HL163640‐03S1; R01HL163640‐02S1; R21AT012430‐01). Carlos Mena‐Hurtado is a consultant for Terumo, Novo Nordisk, and Cook. He receives research grants from Merck, Shockwave, and the National Institutes of Health. Michael G. Nanna: Dr. Nanna reports current research support from the American College of Cardiology Foundation supported by the George F. and Ann Harris Bellows Foundation, the Patient‐Centered Outcomes Research Institute (PCORI), the Yale Claude D. Pepper Older Americans Independence Center (P30AG021342), and the National Institute on Aging (K76AG088428). Dr. Nanna also reports being a consultant for Novo Nordisk, Heartflow, and Merck. Mikhail N. Kosiborod has served as a consultant or on an advisory board for 35Pharma, Alnylam, Amgen, Applied Therapeutics, Arrowhead Pharmaceuticals, AstraZeneca, Bayer, Boehringer Ingelheim, Corcept Therapeutics, Cytokinetics, Dexcom, Eli Lilly, Esperion Therapeutics, Imbria Pharmaceuticals, Janssen, Lexicon Pharmaceuticals, Merck (Diabetes and Cardiovascular), Novo Nordisk, Pfizer, Pharmacosmos, Regeneron, Roche, Sanofi, scPharmaceuticals, Structure Therapeutics, Vifor, and Youngene Therapeutics; has received research grants from AstraZeneca, Boehringer Ingelheim, and Pfizer; holds stock options in Artera Health and Saghmos Therapeutics; has received honoraria from AstraZeneca, Boehringer Ingelheim, and Novo Nordisk; has received other research support from AstraZeneca and Vifor; and is employed by AstraZeneca R&D, effective January 6, 2025.
PEER REVIEW
The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/dom.70080.
Supporting information
Data S1.Supporting Information.
ACKNOWLEDGEMENTS
Medical writing assistance was provided by Shelley Batts, PhD, an independent contractor of Analysis Group, Inc. This study was funded by Novo Nordisk, Inc.
Smolderen KG, Mena‐Hurtado C, Zhao Z, et al. Lower risk of cardiovascular events in patients initiated on semaglutide 2.4 mg in the real‐world: Results from the SCORE study (Semaglutide Effects on Cardiovascular Outcomes in People with Overweight or Obesity in the Real World). Diabetes Obes Metab. 2025;27(11):6691‐6704. doi: 10.1111/dom.70080
The affiliation provided for Mikhail N. Kosiborod is current as of the time the study was conducted.
DATA AVAILABILITY STATEMENT
All data generated or analyzed during the study were derived from the privately held Komodo Research Data and are not publicly available.
REFERENCES
- 1. Henning RJ. Obesity and obesity‐induced inflammatory disease contribute to atherosclerosis: a review of the pathophysiology and treatment of obesity. Am J Cardiovasc Dis. 2021;11(4):504‐529. [PMC free article] [PubMed] [Google Scholar]
- 2. Welsh A, Hammad M, Piña IL, Kulinski J. Obesity and cardiovascular health. Eur J Prev Cardiol. 2024;31(8):1026‐1035. doi: 10.1093/eurjpc/zwae025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Saito Y, Yaginuma H, Asada K, et al. Prognostic impact of previous atherosclerotic cardiovascular disease on short‐ and long‐term outcomes in acute myocardial infarction. Am J Cardiol. 2024;217:18‐24. doi: 10.1016/j.amjcard.2024.02.007 [DOI] [PubMed] [Google Scholar]
- 4. Raisi‐Estabragh Z, Kobo O, Mieres JH, et al. Racial disparities in obesity‐related cardiovascular mortality in the United States: temporal trends from 1999 to 2020. J Am Heart Assoc. 2023;12(18):e028409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. United States Food and Drug Administration . Highlights of Prescribing Information: WEGOVY (Semaglutide). Accessed January 15, 2025. https://www.accessdata.fda.gov/drugsatfda_docs/label/2024/215256s011lbl.pdf
- 6. Lincoff AM, Brown‐Frandsen K, Colhoun HM, et al. Semaglutide and cardiovascular outcomes in obesity without diabetes. N Engl J Med. 2023;389(24):2221‐2232. doi: 10.1056/NEJMoa2307563 [DOI] [PubMed] [Google Scholar]
- 7. Cousins HC, Altman RB. Association between spironolactone use and COVID‐19 outcomes in population‐scale claims data: a retrospective cohort study. medRxiv. Preprint, 2023. doi: 10.1101/2023.02.28.23286515. [DOI] [Google Scholar]
- 8. Gu J, Epland M, Ma X, Park J, Sanchez RJ, Li Y. A machine‐learning algorithm using claims data to identify patients with homozygous familial hypercholesterolemia. Sci Rep. 2024;14(1):8890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Komodo Health Inc . Healthcare Map . Accessed January 15, 2025. https://www.komodohealth.com/technology/healthcare-map
- 10. Kosiborod M, Cavender MA, Fu AZ, et al. Lower risk of heart failure and death in patients initiated on sodium‐glucose cotransporter‐2 inhibitors versus other glucose‐lowering drugs: the CVD‐REAL study (Comparative effectiveness of cardiovascular outcomes in new users of sodium‐glucose cotransporter‐2 inhibitors). Circulation. 2017;136(3):249‐259. doi: 10.1161/circulationaha.117.029190 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Htoo PT, Buse J, Cavender M, et al. Cardiovascular effectiveness of sodium‐glucose cotransporter 2 inhibitors and glucagon‐like peptide‐1 receptor agonists in older patients in routine clinical care with or without history of atherosclerotic cardiovascular diseases or heart failure. J Am Heart Assoc. 2022;11(4):e022376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Hicks KA, Mahaffey KW, Mehran R, et al. 2017 cardiovascular and stroke endpoint definitions for clinical trials. J Am Coll Cardiol. 2018;71(9):1021‐1034. [DOI] [PubMed] [Google Scholar]
- 13. Singh S, Fouayzi H, Anzuoni K, et al. Diagnostic algorithms for cardiovascular death in administrative claims databases: a systematic review. Drug Saf. 2019;42:515‐527. [DOI] [PubMed] [Google Scholar]
- 14. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38(6):1228‐1234. [Google Scholar]
- 15. Aalen OO, Johansen S. An empirical transition matrix for non‐homogeneous Markov chains based on censored observations. Scand J Stat. 1978;5(3):141‐150. [Google Scholar]
- 16. Aalen O. Nonparametric estimation of partial transition probabilities in multiple decrement models. Ann Stat. 1978;6(3):534‐545. [Google Scholar]
- 17. Vickerstaff V, Omar RZ, Ambler G. Methods to adjust for multiple comparisons in the analysis and sample size calculation of randomised controlled trials with multiple primary outcomes. BMC Med Res Methodol. 2019;19(1):129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Plutzky J, Ostrominski JW, Aroda VR, et al. Early clinical benefit of semaglutide in adults with overweight or obesity and cardiovascular disease: a secondary analysis of the SELECT trial (Abstr). Obes Facts. 2025;18(Supp1 1):85. [Google Scholar]
- 19. Ryan DH, Lingvay I, Deanfield J, et al. Long‐term weight loss effects of semaglutide in obesity without diabetes in the SELECT trial. Nat Med. 2024;30(7):2049‐2057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Colhoun H, Lincoff A, Linder M, et al. Exploratory mediation analysis of the effect of semaglutide on cardiovascular outcomes in people with overweight or obesity in the SELECT randomised trial. Eur Heart J. 2024;45(Suppl 1):ehae666.2792. [Google Scholar]
- 21. Holman RR, Bethel MA, Mentz RJ, et al. Effects of once‐weekly exenatide on cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2017;377(13):1228‐1239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Pfeffer MA, Claggett B, Diaz R, et al. Lixisenatide in patients with type 2 diabetes and acute coronary syndrome. N Engl J Med. 2015;373(23):2247‐2257. [DOI] [PubMed] [Google Scholar]
- 23. Ndumele CE, Rangaswami J, Chow SL, et al. Cardiovascular‐kidney‐metabolic health: a presidential advisory from the American Heart Association. Circulation. 2023;148(20):1606‐1635. [DOI] [PubMed] [Google Scholar]
- 24. Patorno E, Pawar A, Franklin JM, et al. Empagliflozin and the risk of heart failure hospitalization in routine clinical care. Circulation. 2019;139(25):2822‐2830. doi: 10.1161/circulationaha.118.039177 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Patorno E, Pawar A, Wexler DJ, et al. Effectiveness and safety of empagliflozin in routine care patients: results from the EMPagliflozin compaRative effectIveness and SafEty (EMPRISE) study. Diabetes Obes Metab. 2022;24(3):442‐454. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1.Supporting Information.
Data Availability Statement
All data generated or analyzed during the study were derived from the privately held Komodo Research Data and are not publicly available.