Recommended and Prevalent Use of Glucagon-like Peptide-1 Receptor Agonists and Sodium–Glucose Cotransporter-2 Inhibitors in a National Population-Based Sample

Background:

The 2022 consensus report from the American Diabetes Association and European Association for the Study of Diabetes recommended using glucagon-like peptide-1 receptor agonists (GLP-1 RAs) among persons with type 2 diabetes (T2D) who have established or are at high risk for atherosclerotic cardiovascular disease (ASCVD), and using sodium–glucose cotransporter-2 inhibitors (SGLT2is) among those with T2D who have established ASCVD, chronic kidney disease, or heart failure or are at high risk for ASCVD (level A evidence) (1).

Objective:

To estimate, using nationally representative data, the number and percentage of persons with T2D who would meet the recommended criteria and were using these medications in the United States.

Methods and Findings:

We included nonpregnant adults aged 20 years or older in NHANES (National Health and Nutrition Examination Survey) during 2017 to 2020 with self-reported diabetes (n = 1417). We excluded those with type 1 diabetes (first diagnosed at age <30 years, started insulin therapy within 1 year of diagnosis, and currently using insulin) (n = 19). Heart failure and ASCVD, including coronary heart disease, angina, myocardial infarction, or stroke, were based on self-report. Persons at high risk for ASCVD were defined as those aged 55 years or older with 2 or more of the following conditions: obesity, hypertension, smoking, dyslipidemia, and albuminuria (1). Chronic kidney disease was defined as a urinary albumin–creatinine ratio of 30 mg/g or higher or an estimated glomerular filtration rate of 20 to 59 mL/min/1.73 m² (1). Medication usage was self-reported. We estimated population counts and percentages, overall and stratified by clinical factors (recommendation criterion, diabetes duration, and body mass index eligibility) and sociodemographic factors (sex, age, race/ethnicity, health insurance, and income level).

After excluding those with incomplete information (n = 68), our study included 1330 participants. We estimated that 22.4 million U.S. adults with diagnosed T2D (82.3%) per year would meet the recommended criteria to use the 2 medications (Table). Most Medicare beneficiaries would be recommended to use them. Of those meeting the criteria, 3.7% used GLP-1 RAs, 5.3% used SGLT2is, and 9.1% used either of them.

Table.

Proportion of U.S. Adults With Diagnosed T2D Who Use and Meet Criteria for GLP-1 RA and SGLT2i Treatment Based on 2022 ADA–EASD Guidelines, 2017–2020 NHANES (n = 1330)

Characteristic	Participants, n	Proportion Who Meet Recommended Criteria (95% CI), %*	Population Count, n (millions)†	Proportion Using GLP-1 RAs or SGLT2is (95% CI), %*
				Who Meet Criteria (n = 1133)	Who Do Not Meet Criteria (n = 197)
All
GLP-1 RAs or SGLT2is‡	1330	82.3 (78.8–85.4)	22.4	9.1 (6.2–13.1)	12.1 (6.1–22.3)
GLP-1 RAs	1320	46.0 (42.0–49.9)	12.5	3.7 (2.2–6.3)	7.2 (2.7–17.9)
SGLT2is	1330	82.3 (78.8–85.4)	22.4	5.3 (3.1–9.0)	4.9 (1.8–12.8)
Clinical factors
By criterion§
ASCVD	294	100	6.6	9.5 (5.3–16.5)	–
Heart failure	157	100	2.7	12.5 (4.7–29.3)	–
CKD	337	100	7.2	8.6 (5.2–13.8)	–
High ASCVD risk	345	100	5.9	7.6 (4.5–12.5)	–
By diabetes duration
0–5 y	385	72.9 (67.3–77.9)	6.2	9.2 (5.1–15.9)	11.9 (4.6–27.3)
6–10 y	266	84.6 (77.4–89.8)	4.9	5.3 (2.7–10.0)	13.6 (6.7–25.9)
≥11 y	662	87.5 (81.6–91.7)	11.3	10.6 (6.7–16.2)	11.2 (2.1–42.8)
By BMI eligibility‖
Below BMI threshold	251	83.0 (75.1–88.7)	4.1	5.3 (2.1–12.5)	6.1 (1.8–18.1)
Met BMI threshold	984	82.3 (77.7–86.2)	18.3	10.0 (7.0–14.1)	13.9 (6.6–26.7)
Sociodemographic factors
By sex
Female	612	79.5 (73.8–84.3)	9.8	6.4 (3.7–10.8)	8.1 (3.9–16.0)
Male	718	84.6 (79.9–88.4)	12.6	11.1 (6.5–18.4)	16.4 (6.5–35.5)
By age
<65 y	668	70.0 (63.4–75.8)	10.3	11.5 (7.3–17.5)	13.0 (6.5–24.1)
≥65 y	662	97.0 (95.0–98.2)	12.1	7.0 (3.9–12.4)	1.4 (0.2–11.1)
By race
White	425	83.6 (77.9–88.0)	13.6	9.9 (5.6–16.8)	16.5 (6.9–34.6)
Non-White	905	80.5 (76.9–83.7)	8.9	7.9 (6.1–10.0)	6.6 (3.2–13.2)
By health insurance
Private	278	65.4 (55.3–74.4)	5.4	14.8 (7.8–26.5)	13.7 (6.6–26.1)
Medicare	748	94.5 (90.9–96.7)	13.2	7.7 (4.7–12.3)	19.6 (2.9–66.5)
Medicaid	98	77.5 (64.5–86.7)	1.2	11.3 (4.5–25.8)	8.3 (1.5–35.2)
Other or no insurance	200	74.8 (63.2–83.7)	2.6	3.9 (0.7–18.0)	2.1 (0.3–15.7)
By poverty¶
Above	796	83.2 (78.4–87.2)	19.2	9.4 (6.1–14.2)	10.1 (4.4–21.4)
Below	251	83.8 (76.7–89.1)	3.5	9.8 (5.1–17.9)	5.4 (1.2–21.7)

ADA = American Diabetes Association; ASCVD = atherosclerotic cardiovascular disease; BMI = body mass index; CKD = chronic kidney disease; EASD = European Association for the Study of Diabetes; GLP-1 RA = glucagon-like peptide-1 receptor agonist; NHANES = National Health and Nutrition Examination Survey; SGLT2i = sodium–glucose cotransporter-2 inhibitor; T2D = type 2 diabetes.

^*We accounted for the complex, multistage, and probability sampling design of NHANES and used the mobile examination center examination weights in our analyses to produce national estimates that represent the noninstitutionalized civilian population in the United States.

^†Estimated by multiplying the weighted prevalence in subgroups (e.g., all, by criterion, and by diabetes duration) by the U.S. civilian noninstitutionalized population count of adults aged ≥20 y. Due to rounding, the sum of the population count from each subgroup may not be exactly equal to the total count.

^‡The medications were categorized into therapeutic classes using the Multum MediSource Lexicon classification system (Cerner). Using the drug classification codes, we identified the GLP-1 RA medications exenatide, liraglutide, albiglutide, dulaglutide, and semaglutide and the SGLT2i medications dapagliflozin, canagliflozin, empagliflozin, and ertugliflozin.

^§We used the following mutually exclusive categorization of the criteria: ASCVD without heart failure, heart failure, CKD without ASCVD and heart failure, high ASCVD risk without ASCVD, heart failure, and CKD.

^‖Defined as BMI ≥25 kg/m² for Asian Americans and BMI ≥27 kg/m² for all other races. BMI data from NHANES were calculated on the basis of measured height and weight.

^¶Above poverty is defined as the ratio of reported family monthly income to federal poverty level being >1, and below poverty is defined as the ratio being ≤1.

Discussion:

On the basis of nationally representative data, we estimated that many U.S. patients with T2D would meet the recommended criteria to use GLP-1 RAs or SGLT2is. Only 1 in 10 of them used either of these medications between 2017 and 2020, a period when these 2 medications were not recommended as first-line therapy to many patients now eligible in the 2022 guidelines.

Our estimated percentages of patients who would meet the recommended criteria were higher than previous estimates (2), which may be attributable to our including persons with high risk for ASCVD for recommended use of an SGLT2i in the analysis and using newer data. Our estimated percentages of recommended patients with T2D who used these medications are lower than the results (8% for GLP-1 RAs and 11.2% for SGLT2is) noted in a prior study (3) of Veterans Affairs patients.

Clinical studies have shown that both GLP-1 RAs and SGLT2is yield additional clinical benefits compared with older treatments in reducing body weight and progression of CVD and chronic kidney disease. However, these medications come at a substantially higher cost. Data suggest that these 2 medications are cost-effective as second-line agents (4). However, the current cost would have to decrease by 70% for these to be cost-effective as first-line agents (5). Further studies are needed to assess if these newer medications are cost-effective among particular patient subgroups as first-line agents.

Our study is subject to several limitations. First, some health conditions are self-reported. Second, our data predate the most recent guidelines. Third, use of these medications may have increased since 2020.

More than 80% of U.S. adults and nearly all Medicare beneficiaries with T2D would meet the criteria to use either GLP-1 RAs or SGLT2is based on the recommendations from the American Diabetes Association and European Association for the Study of Diabetes. However, at current drug pricing, using these 2 new medications as first-line agents among all eligible patients with T2D may not be cost-effective. Cost-effectiveness was not formally considered in the current guideline, but an assessment of cost-effectiveness may assist better targeting of interventions to achieve the greatest effect (1) at a sustainable cost.