The identification of patients eligible for the analysis by year is shown in online supplemental figure S1. The number of eligible patients ranged from 279,474 to 348,997 between 2018 and 2022.

The percentage of the prevalent use of a GLP-1 RA and/or SGLT2i prescription increased almost threefold from 9.2% in 2018 to 27.1% in 2022 (figure 2). The proportion of patients with a prescription for a GLP-1 RA alone rose 1.9-fold from 5.2% in 2018 to 9.9% in 2022. For SGLT2i alone, the proportion rose 4.4-fold from 2.8% in 2018 to 12.2% in 2022. The proportion of patients with a prescription for both a GLP-1 RA and an SGLT2i increased from 1.2% in 2018 to 5.0% in 2022.

Cross-sectional study outcomes. Proportion of patients with T2D and ASCVD who were prescribed GLP-1 RAs or SGLT2is by year. ASCVD, atherosclerotic cardiovascular disease; GLP-1 RA, glucagon-like peptide-1 receptor agonist; SGLT2i, sodium-glucose cotransporter 2 inhibitor; T2D, type 2 diabetes.

The attrition of patients identified for inclusion in the nested cohort study (evaluating the proportion of patients with T2D with a new diagnosis of ASCVD who were newly prescribed a GLP-1 RA or SGLT2i) is presented in online supplemental figure S2. The study cohort comprised a total of 203,115 patients, for whom demographic and clinical characteristics are summarized in online supplemental table S1. The mean (SD) age of this cohort was 64.7 (13.0) years; 53.4% were women, and a majority (67.8%) were White in race. The most represented US region was the South (44.4%) followed by the Northeast (29.2%), Midwest (15.6%), and the West (9.7%).

Data concerning the morbidities and medications of the cohort are also summarized in online supplemental table S1. The most common comorbidities included hypertension (70.6%), dyslipidemia (56.9%), and nephropathy (25.2%). The most common types of ASCVD diagnoses at the index included other coronary heart disease (45.8%), peripheral artery disease (29.7%), and ischemic stroke (15.5%). The most frequently used other antidiabetic medications at baseline were an insulin regimen, including basal and other insulin (41.5%), biguanides (29.9%) and sulfonylureas (12.0%).

Across the entire study period, 16,635 patients (8.2%) were newly prescribed a GLP-1 RA and/or SGLT2i in the year following a diagnosis of ASCVD (figure 3). However, the percentage of incident use of these agents (GLP-1 RA and/or SGLT2i) increased almost threefold between 2018 and 2022, from 5.9% to 17.0%.

Nested cohort outcomes. Proportion of patients with T2D and newly diagnosed ASCVD who were newly prescribed GLP-1 RAs or SGLT2is, overall and by year. ASCVD, atherosclerotic cardiovascular disease; GLP-1 RA, glucagon-like peptide-1 receptor agonist; SGLT2i, sodium-glucose cotransporter 2 inhibitor; T2D, type 2 diabetes.

Considering each class separately, GLP-1 RAs only and SGLT2is only were prescribed to 8,388 (4.1%) and 6,481 (3.2%) patients, respectively, across the study period (figure 3). In 2018, the proportion of patients given a GLP-1 RA prescription was 3.6%, compared with 1.8% of patients given an SGLT2i prescription. In 2022, these proportions had increased by 2.2-fold to 7.8%, and 3.9-fold to 7.0%, respectively. The number of patients given both a GLP-1 RA and SGLT2i across the study period was 1,766 (0.9%). This proportion also increased each year, from 0.5% in 2018 to 2.3% in 2022.

Some demographic differences were observed when comparing the 8.2% of the study cohort who were newly prescribed a GLP-1 RA and/or SGLT2i with the 91.8% who were not (online supplemental table S1). Patients with a prescription for a GLP-1 RA and/or SGLT2i were generally younger than patients without (mean age 62.2 years for GLP-1 RA and/or SGLT2i (ASD 22.5%), 61.0 years for GLP-1 RA only (ASD 32.0%), 64.3 years for SGLT2i only (ASD 5.6%), 60.3 years for both GLP-1 RA and SGLT2i (ASD 38.7%), compared with 65.0 years for no GLP-1 RA and/or SGLT2i prescription). This was likely driven by the greater proportion of patients aged ≥65 years in the group without a prescription for a GLP-1 RA and/or SGLT2i than those with a prescription (57.9% vs 45.2%, ASD 25.5%; online supplemental table S1).

In the Midwest region of the USA, a lower proportion of people had a prescription for a GLP-1 RA and/or SGLT2i agent compared with those who did not (12.3% vs 15.9%, ASD 10.4%). Conversely, in the Northeast region, a greater proportion of people had a prescription for a GLP-1 RA and/or SGLT2i agent compared with those who did not (34.6% vs 28.7%, ASD 12.8%). No meaningful differences were observed in the South or West regions of the USA (online supplemental table S1).

Patients with a prescription for a GLP-1 RA and/or SGLT2i were also observed to be receiving a higher mean number of antidiabetic medications at baseline (1.5 for GLP-1 RA and/or SGLT2i (ASD 48.5%), 1.5 for GLP-1 RA only (ASD 53.7%), 1.4 for SGLT2i only (ASD 42.8%), 1.4 for both GLP-1 RA and SGLT2i (ASD 44.7%) compared with 0.9 for no GLP-1 RA and/or SGLT2i prescription; online supplemental table S1). A greater proportion of patients with a prescription for a GLP-1 RA alone were also prescribed basal/other insulin, biguanides, sulfonylureas, thiazolidinediones, and DPP4is when compared with patients without a prescription for a GLP-1 RA or SGLT2i (online supplemental table S1). Similar observations were made for patients with a prescription for an SGLT2i alone, except for “other insulins”, of which the proportion was similar to patients not prescribed with a GLP-1 RA or SGLT2i.

A greater proportion of patients with a prescription for a GLP-1 RA and/or SGLT2i had metabolic disease at baseline, including hyperosmolarity, ketoacidosis, and hypoglycemia with/without coma (22.8% vs 16.8%, ASD 15.0%) and neuropathy (25.4% vs 20.9%, ASD 10.6%) when compared with patients without a prescription. This discrepancy was greatest between patients prescribed a GLP-1 RA alone and patients without a prescription for either drug class (online supplemental table S1). Conversely, patients who were prescribed a GLP-1 RA only and both a GLP-1 RA and SGLT2i had lower proportions of baseline CVD (19.6% vs 24.0% (ASD 10.8%) and 17.6% vs 24.0% (ASD 16.0%), respectively) and baseline cancer (11.9% vs 16.4% (ASD 12.9%), and 10.0% vs 16.4% (ASD 16.4%), respectively) compared with patients who were not prescribed GLP-1 RAs and/or SGLT2is. Patients who were prescribed SGLT2is only and both a GLP-1 RA and SGLT2i had lower proportions of baseline nephropathy (21.0% vs 25.4% (ASD 10.4%), and 20.2% vs 25.4% (ASD 12.4%), respectively) and baseline anxiety (14.5% vs 20.8% (ASD 16.7%), and 16.4% vs 20.8% (ASD 11.5%), respectively) compared with patients who were not prescribed GLP-1 RAs and/or SGLT2is. Another observation considered as a meaningful difference was the greater proportion of patients with HF (13.5% vs 9.8%, ASD 11.3%) and smaller proportion of patients with depression (15.0% vs 20.0%, ASD 13.1%) among those prescribed an SGLT2i when compared with patients without a prescription for GLP-1 RA/SGLT2i (9.8%). People prescribed SGLT2is-only or both GLP-1 RA and SGLT2i had a lower burden of comorbidities as evidenced by a lower baseline Charlson Comorbidity Index score compared with people without a prescription for GLP-1 RA/SGLT2i. People prescribed GLP-1 RA or SGLT2i or GLP-1 RAs-only had the greatest differences in baseline antidiabetic drug and ASCVD-related drug burden compared with those without GLP-1 RA/SGLT2i (online supplemental table S1).

Higher percentages of baseline myocardial infarction, ischemic stroke, other coronary heart disease, or peripheral artery disease were found between those prescribed SGLT2is only and those without a prescription for GLP-1 RA or SGLT2i. Additionally, greater proportions of myocardial infarction and other coronary heart disease at baseline were observed in those prescribed both GLP-1 RA and SGLT2i compared with those without a prescription for GLP-1 RA or SGLT2i (online supplemental table S1).