Use of Second-Generation Antidiabetic Medication Among a Nationally Representative Sample

ABSTRACT

Objectives: Existing studies have shown the benefits of second-generation antidiabetic medications in patients with type 2 diabetes (T2D). However, the medications’ real-world utilization was not well understood. Our study assessed patient factors associated with the use of second-generation antidiabetic medications in a nationally representative sample of patients with T2D.

Study Design: This retrospective, cross-sectional analysis used the 2005 to 2018 National Health and Nutrition Examination Survey (NHANES) data.

Methods: Survey participants 18 years and older who had a diagnosis of T2D and had used antidiabetic medications in the past 30 days were included. The primary outcome was the prescription of any second-generation antidiabetic medication. Weighted stepwise multivariable logistic regression models were used to assess the associations between the use of second-generation antidiabetic medications and patients’ characteristics.

Results: Among 4493 patients with T2D, 533 (weighted %, 13.67%) reported using at least 1 second-generation antidiabetic drug. In multivariable analyses, patients with incomes at least 400% of the federal poverty level (adjusted odds ratio [AOR], 2.30; 95% CI, 1.58-3.34), with higher hemoglobin A_1c levels (AOR, 1.10; 95% CI, 1.02-1.18), and taking more medications (AOR, 1.14; 95% CI, 1.09-1.20) were more likely to use second-generation antidiabetic drugs compared with their counterparts.

Conclusions: The uptake of second-generation antidiabetic medications was 14% among patients with T2D in the United States. Prescription benefit design that targets lower out-of-pocket payments for these newer drugs may improve patient access and clinical outcomes for patients with T2D.

Am J Manag Care. 2023;29(10):e307-e316. https://doi.org/10.37765/ajmc.2023.89445

Takeaway Points

This study aims to assess the use of newer drugs for the treatment of patients with type 2 diabetes in the United States. We found that the use of newer drugs was 14%. In addition, newer drugs were prescribed more frequently to patients with higher income, worse glucose control, and more comorbidities. Our findings are expected to improve awareness of barriers to patient uptake of antihyperglycemic medications and access to these newer, expensive treatments in patients with type 2 diabetes.

• Nationally, approximately 14% of patients with type 2 diabetes receiving antidiabetic drugs have used at least 1 second-generation antidiabetic medication. Dipeptidyl peptidase 4 inhibitors were the most frequently prescribed newer agents.
• Income level, hemoglobin A_1c, and comorbidity burden were the independent patient-level drivers of the use of newer agents.

In the United States, type 2 diabetes (T2D) affects approximately 34.2 million individuals, or approximately 10% of the population,¹ with 1.5 million new cases diagnosed each year.² T2D accounts for 90% to 92% of total diabetes diagnoses.¹ The estimate of health care expenditure for diabetes was $327 billion in 2017, marking a 26% increase from 2012 ($245 billion), including $237 billion in direct medical costs and $90 billion in productivity loss.³ On an individual level, the mean medical cost for patients with diabetes is $16,752, of which 57% is from diabetes care and 15% from antidiabetic medications.³

Diabetes can be managed, and its consequences can be avoided or delayed with diet, physical activity, medication, and regular screening and treatment for complications.⁴ The American Diabetes Association (ADA) in 2012 started to recommend a shared decision-making approach to individualize glycemic targets and medication selection based on clinical characteristics, socioeconomic status, and patient preference.⁵ According to the latest ADA guidelines, metformin is recommended for the initial pharmacologic treatment of patients with T2D.⁶ In the early 2000s, the treatment options for T2D, in addition to metformin, were limited to sulfonylureas, thiazolidinediones, and insulin. Although these antidiabetic agents have shown potent antihyperglycemic effects, their major disadvantages are elevated risks of hypoglycemia, weight gain, bone fracture, and heart failure.⁶

Since 2005, the FDA has approved glucagonlike peptide-1 (GLP-1) receptor agonists (2005),⁷ dipeptidyl peptidase 4 (DPP-4) inhibitors (2006),⁸ and sodium-glucose cotransporter 2 (SGLT2) inhibitors (2013).⁹ These second-generation antidiabetic medications have less hypoglycemic effect and weight gain but cost more than traditional agents.⁶ For example, the generic forms of first-generation antihyperglycemic medications are available for $4 per month, whereas second-generation medications may cost more than $500 per month.^10-12 DeJong et al found that annual costs for common novel agents ranged from $5202 to $11,225 compared with $31 to $136 for traditional drugs in Medicare Part D.¹³ Liss et al also found that the higher cost of nonsulfonylurea medications (ie, DPP-4 inhibitors, GLP-1 receptor agonists, and SGLT2 inhibitors) was the main driver of relative increases in total costs in a large cohort of commercially insured adults with T2D in the United States.¹⁴ In addition, studies have shown the benefits of second-generation antidiabetic medications beyond their antihyperglycemic effect, such as preventing atherosclerotic cardiovascular disease (ASCVD; SGLT2 inhibitors and GLP-1 receptor agonists), preventing heart failure (SGLT2 inhibitors), and improving renal function (SGLT2 inhibitors and GLP-1 receptor agonists).⁶ Given the proven benefits of these agents, their uptake was expected to increase since 2015. However, the main factor that has hindered the uptake of second-generation antidiabetic prescriptions and access to these new medications is cost.¹⁵ In addition, the decision to prescribe these new medications in clinical practice can be complicated by other factors such as patients’ demographics, number of comorbidities, and the presence of specific comorbidities.⁶

Results of recent studies have updated the national utilization of antidiabetic medications. Gilstrap et al found that among Medicare beneficiaries from 2007 to 2015, 7% of patients with newly diagnosed T2D initiated second-generation antidiabetic agents.¹⁶ In this study, the uptakes of DPP-4 inhibitors and GLP-1 receptor agonists were associated with patients’ comorbidities such as chronic kidney disease (CKD) and ischemic heart disease, respectively.¹⁶ In 2 other studies, the authors explored the use of all therapeutic classes for the treatment of T2D and whether the real-world prescribing reflected the recommendations of ADA in a national representative sample.^17,18 Specifically, they explored adherence to the ADA guidelines such as metformin as first-line treatment and medication selection based on risk of hypoglycemia, weight status, cost, and CV benefits.¹⁷ However, treatment selection for patients with obesity or CVD was not consistent with ADA recommendations.¹⁷ In addition, the associations between patient factors (eg, demographics, socioeconomic status, health behavior, clinical characteristics of diabetes, comorbidities) and the use of second-generation antidiabetic agents remain unknown. Understanding these associations will help clinical practitioners and policy makers tailor treatment choices and develop initiatives to enhance the uptake of second-generation antidiabetic agents.

This study assessed patient factors associated with the use of second-generation antidiabetic medications in a nationally representative sample of patients with T2D. We hypothesized that patients with a high risk of hypoglycemia (ie, older adults and those with activities of daily living [ADL] dependency, who have more comorbidities, and who take more medications) are more likely to receive the second-generation antidiabetic agents.⁶ In addition, patients with the intention of weight loss (ie, those with obesity or on a diabetic diet), specific comorbidities (eg, CVD, CKD), and higher socioeconomic status may have a higher likelihood of using a second-generation antidiabetic agent.⁶

METHODS

Study Design, Data Sources, and Study Sample

We conducted a retrospective, cross-sectional analysis using the 2005 to 2018 National Health and Nutrition Examination Survey (NHANES) data. The NHANES is a nationally representative survey that uses a complex sampling design to collect deidentified information on individuals’ demographics, health behaviors, nutrition intake, medication use, physical and laboratory examinations, diseases, and conditions among the noninstitutionalized US population through in-home interviews and mobile examination centers across the United States.¹⁹ Compared with payers’ administrative claims data, the NHANES includes additional data on individuals’ socioeconomic information, health behavior and lifestyle, laboratory tests, and health-related quality of life. This study was reviewed and exempted by the Auburn University Institutional Review Board.

The study sample included NHANES adult participants (aged ≥ 18 years) who either had a diagnosis of diabetes by ADA criteria or had been told they have diabetes.²⁰ Participants with diabetes were identified if any of the following criteria were met, using the NHANES laboratory tables: (1) hemoglobin A_1c (HbA_1c) of 6.5% or greater, (2) a 2-hour plasma glucose of at least 200 mg/dL (11.1 mmol/L) during oral glucose tolerance test, (3) fasting plasma glucose of at least 126 mg/dL (7.0 mmol/L), or (4) had ever been told they have diabetes or reported taking any antidiabetic pills in the NHANES Diabetes Questionnaire. The study sample was further limited to patients with diabetes who had taken any antidiabetic medication in the past 30 days identified from the NHANES Prescription Drugs table, using Multum MediSource Lexicon’s therapeutic category name (second-level category name = 99). We then removed patients with potential type 1 diabetes by excluding patients younger than 20 years and using only insulin, as well as pregnant women.²¹ In addition, participants with nonpositive sampling weight were excluded because they did not contribute to the weighted analysis. All laboratory results in the NHANES were obtained using certified and standardized procedures.²²

Measurements

Outcome. The primary outcome was the prescription of any second-generation antidiabetic medication. Patients were recorded as taking any second-generation antidiabetic medication if their prescription history in the past 30 days contained any component of GLP-1 receptor agonists, DPP-4 inhibitors, and SGLT2 inhibitors identified by Multum MediSource Lexicon’s therapeutic category code, regardless of single-ingredient or combination products (eAppendix Table [available at ajmc.com]). Otherwise, patients were categorized into the subgroup of those who did not use second-generation antidiabetic medication.

Patient factors. Patient factors assessed in this study included patients’ demographics, socioeconomics, access to care, health behaviors, diabetes-related variables, and comorbidities that may be associated with the use of second-generation antidiabetic medications. Details in these 6 domains of patient factors are described as follows:

• Demographics (Demographic Variables & Sample Weights table in NHANES): age group (18-44, 45-64, or ≥ 65 years), sex (male or female), and race/ethnicity (Hispanic, non-Hispanic Black, non-Hispanic White, or other).
• Socioeconomic status (Demographic Variables & Sample Weights table): education status (less than high school, high school, or college and above), marital status (widowed/divorced/separated, married/living with a partner, or never married), and poverty level (defined by the ratio of income to federal poverty level [FPL]: < 100%, 100%-<200%, 200%-<400%, or ≥ 400%).
• Access to care (Health Insurance table): health insurance status (uninsured vs any health insurance) and prescription coverage status (yes vs no).
• Health behaviors: Alcohol use was defined as nonuse, moderate use (≤ 2 drinks/day for men and ≤ 1 drink/day for women), or heavy use (> 2 drinks/day for men and > 1 drink/day for women) from the Alcohol table in NHANES, according to Dietary Guidelines for Americans 2020-2025.²³ Smoking status was categorized as never smoked, current smoker, or former smoker (Smoking—Cigarette Use table). Physical activity levels were defined as tertiles of total metabolic equivalent scores calculated by an algorithm provided for each cycle: low, moderate, or rigorous (Physical Activity table).²⁴ Body mass index was classified as less than 30 kg/m² (underweight, normal, overweight) or at least 30 kg/m² (obese) from the Body Measures table.²⁵ ADL dependency was defined by any difficulty in any of the 3 components of ADL in the physical functioning questionnaire: getting in and out of bed, eating, and dressing (Physical Functioning table).
• Diabetes-related factors: HbA_1c at the time of medical examination (continuous variable; Glycohemoglobin table), current diabetic diet (yes if the patient is on any diabetic diet; Dietary Interview—Total Nutrient Intakes table), and the presence of any diabetic complication (if the patient reported that their diabetes affected their eyes or they had retinopathy; Diabetes table).
• Comorbidities and medication use: We obtained 16 continuously reported comorbidities from the Medical Conditions table (2005-2018). Specific comorbidities that may be related to second-generation antidiabetic medication are any CVD (ie, angina, stroke, myocardial infarction, coronary heart disease), heart failure (HF), and CKD. CKD was defined by an estimated glomerular filtration rate (eGFR) of less than 60 mL/min/1.73 m² or albuminuria (Albumin & Creatinine—Urine table).²⁶ eGFR was calculated based on the Chronic Kidney Disease Epidemiology Collaboration creatinine equation and the Kidney Disease: Improving Global Outcomes definition with provided serum creatinine level and race of participants.^26,27 Serum creatinine was calibrated across cycles.²⁸ Albuminuria was defined as an albumin-to-creatinine ratio of at least 30 mg/g.²⁶ Number of comorbidities and number of medications taken in the past 30 days were also obtained (Prescription Medications table).

Statistical Analysis

Descriptive statistics (continuous variables: mean and SD; categorical variables: frequency count and percentage) were used to describe the study sample. Bivariable analyses were conducted to compare the differences in distributions of assessed patient factors between patients who used and did not use any second-generation antidiabetic medications. We used domain-based multivariable logistic regression models to assess the associations between patient factors and the use of any second-generation antidiabetic medications. Specifically, we conducted stepwise logistic regression modeling by adding each domain of patient factors into the multivariable models and assessing the fluctuation of effect estimate (adjusted odds ratio [AOR] with 95% CI) with an additional domain. Health insurance status was not included in the model due to its high correlation with prescription coverage, and prescription coverage may be a better predictor for the use of second-generation medications in practice. Likewise, we selected number of medications taken and dropped number of comorbidities out of the model due to correlation. We also tested the interactions between race/ethnicity and prescription coverage using the F test. Sensitivity analyses were conducted by stratifying the prescriptions of second-generation antidiabetic medications into GLP-1 receptor agonists, DPP-4 inhibitors, and SGLT2 inhibitors as well as time before 2015 and after 2015 (when the evidence on beyond-antihyperglycemic effects of SGLT2 inhibitors and GLP-1 receptor agonists became available). All analyses were weighted using the 2-year cycle NHANES weights (interview weights and medical exam weights–Mobile Examination Center [MEC]).²⁹ Weights were adjusted for a combined 7 cycles (2005-2018).³⁰ According to the NHANES analytic guide, we used MEC because it yielded the least common denominator (the smallest subpopulation that includes all the variables).³⁰ Patient observations with missing data under some variables were removed from the multivariable analysis if the missing proportion was less than 1% in those variables and were classified as unknown to replace missing values if otherwise. Statistical analyses were conducted using SAS 9.4 (SAS Institute) with the significance level being 0.05.

RESULTS

Study Sample and Patients’ Characteristics

The study sample included 4493 patients with T2D treated with antidiabetic medications, which represented 19,629,829 patients across the United States from 2005 to 2018 (Figure). Among them, 533 (weighted proportion, 13.67%) reported the use of at least 1 second-generation antidiabetic drug in the past 30 days. Specifically, 100 patients (weighted, 21.6%; 95% CI, 16.67%-26.71%) used GLP-1 receptor agonists, 400 patients (weighted, 70.62%; 95% CI, 64.70%-76.54%) used DPP-4 inhibitors, and 57 patients (weighted, 12.34%; 95% CI, 7.53%-17.16%) used SGLT2 inhibitors.

Characteristics of patients with and without the use of second-generation antidiabetic medications are described in Table 1 [part A and part B]. Since the drugs’ launch into the US market in 2005, the proportion of users has increased consistently across years.

Patient Factors Associated With the Use of Second-Generation Antidiabetic Medications (unadjusted)

For age, race/ethnicity, poverty level, health insurance, and prescription coverage status, the corresponding largest groups of patients in the study sample who reported taking second-generation antidiabetic medications were those who were aged 45 to 64 years, were non-Hispanic White, had incomes above 400% of the FPL, were insured, and had prescription coverage (all P < .05) (Table 1). In addition, patients who used second-generation antidiabetic medications had higher mean HbA_1c levels and had taken more medications compared with nonusers (all P < .05).

Patient Factors Associated With the Use of Second-Generation Antidiabetic Medications (adjusted)

In the weighted stepwise logistics multivariable models (n = 4446) (Table 2), patients with T2D with the following factors were consistently and significantly associated with receiving second-generation antidiabetic medications in the full model: income of 200% to 400% of the FPL (AOR, 1.69; 95% CI, 1.10-2.59), income of 400% or more of the FPL (AOR, 2.30; 95% CI, 1.58-3.34), higher HbA_1c level (a 1% increase in HbA_1c was associated with a 10.1% increase in odds of receiving newer-generation medications compared with not receiving them; AOR, 1.10; 95% CI, 1.03-1.18), and number of medications taken (AOR, 1.14; 95% CI, 1.09-1.20). Prescription coverage was a significant predictor when health behavior factors were added to the model (AOR, 1.49; 95% CI, 1.06-2.09) but was not significant in the full model (AOR, 1.30; 95% CI, 0.89-1.90). In contrast, patients with current smoking status were less likely to use second-generation antidiabetic agents in the full model (AOR, 0.51; 95% CI, 0.28-0.94), but the association was not significant in the initial model (AOR, 0.61; 95% CI, 0.35-1.04). The interaction between race/ethnicity and prescription coverage was not significant (P = .47).

Sensitivity Analyses

In the first sensitivity analysis, we assessed the associations between patient factors and the use of individual second-generation antidiabetic agents. Patients’ income level being 400% or more of the FPL and their number of medications were consistently and significantly associated with the use of GLP-1 receptor agonists, DPP-4 inhibitors, and SGLT2 inhibitors. However, higher HbA_1c level was associated with the use of GLP-1 receptor agonists but not DPP-4 inhibitors or SGLT2 inhibitors (Table 3). In the second sensitivity analysis, we restricted the study sample to NHANES cycles from 2015 to 2018 when the beyond-hyperglycemic benefits of second-generation antidiabetic agents were published. The findings were consistent with the main analysis, and patients aged 45 to 64 years were more likely to use second-generation antidiabetic agents compared with those aged 18 to 44 years (Table 3).

DISCUSSION

From 2005 to 2018, approximately 14% of patients with T2D treated with antidiabetic medications across the United States received at least 1 second-generation antidiabetic drug. The most commonly prescribed second-generation antidiabetic drug class was DPP-4 inhibitors. Higher income level, inadequate glycemic control (ie, higher HbA_1c level), and a greater number of medications were the main patient factors associated with the use of second-generation antidiabetic agents.

In a recent study, Gilstrap et al found that between 2007 and 2015 only 7% of Medicare beneficiaries received second-generation antidiabetic agents among treated patients with T2D.¹⁶ The main difference in the uptake of these medications between the current study results and those of Gilstrap et al may be attributable to patients with different age groups. Our study included nationally representative adults 18 years and older, with patients 65 years and older accounting for only 36.26% of patients among those who used second-generation antidiabetic agents. However, due to similar study duration period, DPP-4 inhibitors were the most prescribed second-generation antidiabetic agents in both studies. Although a GLP-1 receptor agonist (exenatide) was approved and available on the US market before a DPP-4 inhibitor, the utilization rate of DPP-4 inhibitors was higher. This could be explained by the availability of oral dosage forms of DPP-4 inhibitors and the hesitancy of using GLP-1 receptor agonists as injectable drugs.¹⁶ Because the first SGLT2 inhibitor (canagliflozin) was approved in 2013, the small proportion of SGLT2 inhibitor users found in this study was expected. Le et al also found a small uptake of SGLT2 inhibitors in NHANES cycles from 2013 to 2016.¹⁷

Additionally, income level was associated with the use of second-generation antidiabetic agents among patients with T2D receiving antidiabetic medication treatments. Indeed, second-generation antidiabetic drugs are costly and dominated by brand-name products on the market, whereas traditional/first-generation agents (ie, metformin, sulfonylureas) are available as generic formulations and mostly covered by health insurance. Given the significant difference in co-pay between brand and generic antidiabetic drugs, patients with higher incomes are more likely to be able to afford expensive, brand-name, newer antidiabetic drugs than those with lower incomes. Our findings supported that patients’ income and affordability were important factors in access to second-generation antidiabetic drugs when out-of-pocket cost is a barrier.

In addition, patients with higher HbA_1c levels, which indicate inadequate glycemic control, had greater odds of receiving second-generation antidiabetic drugs. This finding is consistent with the ADA recommendation of combination therapy for patients whose glucose levels are not controlled by first-line treatment such as metformin. Specifically, ADA recommends combination therapy for patients with HbA_1c levels of 1.5% to 2.0% above target.⁶ The selection of second-generation antidiabetic agents also depends on clinical characteristics (such as high ASCVD risk, HF, CKD, other comorbidities, and risk for specific adverse effects) and patients’ preference. However, we could not find significant associations between individual comorbidities such as CVD, HF, and CKD with the use of second-generation antidiabetic medications, overall or for individual drug classes. Our findings might be biased toward the null due to the small number of patients who had these comorbidities in our sample or indicate consistency with previous research showing that antidiabetic treatment selection for patients was not consistent with ADA recommendations.¹⁷ Future research is needed to understand how practitioners’ prescribing behaviors and patients’ preference affect the utilization of second-generation antidiabetic medications.

Furthermore, we found that a greater number of medications taken by patients was associated with higher likelihood of use of second-generation antidiabetic agents. Number of medications can be a proxy of disease burden and illness severity.^31-33 Therefore, our findings imply that patients with a higher number of or more severe conditions (ie, diabetes- or non–diabetes-related conditions) were more likely to receive second-generation antidiabetic agents. In fact, patients with higher disease burden of comorbidities were at higher risk of hypoglycemia.^34-36 Because second-generation antidiabetic agents have minimal hypoglycemic effects,⁶ these medications could be prescribed to patients with a higher risk of hypoglycemia, such as those with higher burden of comorbidities. Although the association between the use of second-generation antidiabetic medications and weight status was not significant in our study, overall and by individual drug class, this might be due to the domination of DPP-4 inhibitor use in our sample; this class is known for its weight-neutral effect.³⁷ ADA recommends second-generation antidiabetic treatments for patients with compelling need to minimize weight gain or induce weight loss.⁶ Weight loss is the first beyond-antihyperglycemic effect of GLP-1 receptor agonists, and the FDA approved liraglutide in 2014 for the treatment of obesity.^38,39 In 2021, semaglutide, another GLP-1 receptor agonist, was approved for obesity treatment.⁴⁰ In addition, the weight-loss effect of SGLT2 inhibitors was found in results from some of the very first randomized clinical trials on this drug class.⁴¹ Although the existing evidence supports the use of GLP-1 receptor agonists and SGLT2 inhibitors for patients with obesity, our study lacks statistical power to detect the association. Future studies may use the NHANES data to assess the association between weight status/weight loss–related activities and the use of such medications when newer data become available.

Limitations

Our study has several limitations. First, cross-sectional and observational study design could assess only association rather than causality. Second, unobserved confounding was inevitable (ie, patient and physician preference, conditions related to secondary hyperglycemia), although the richness of the NHANES data provided access to patient factors including laboratory data and health behaviors, which are often missing in administrative claims data. We implemented a stepwise modeling approach, tested for interaction, and conducted sensitivity analyses to confirm the robustness of main findings. Third, due to the small sample sizes of individual second-generation antidiabetic medication subgroups, we may not have enough statistical power to detect associations between medication use and different patient factors. Therefore, future research in assessing the associations with use of individual second-generation antidiabetic medications is critical. Fourth, selection bias may have occurred because we could not ensure that our exclusion criteria removed all patients diagnosed with type 1 diabetes. Finally, recall bias may persist because some of the patient factors were collected from self-reported data.

However, to our knowledge, this is the first study to evaluate the associations between patient factors and second-generation antidiabetic medications among a nationally representative US sample. The NHANES data capture additional, rich patient-level variables such as socioeconomic information, health behaviors including diet and lifestyle, laboratory tests, and health-related quality of life compared with administrative claims data. Our findings will assist practitioners and policy makers to tailor antidiabetic treatment and design prescription benefit coverage to improve access to these newer, expensive treatments.

CONCLUSIONS

The uptake of second-generation antidiabetic medications was 14% among patients with T2D in the United States. Patients’ income, HbA_1c level, and number of medications taken were significantly associated with the use of second-generation antidiabetic medications. Future research to explore practitioners’ prescribing behaviors and patients’ preferences around second-generation antidiabetic medication prescription is warranted.

Author Affiliations: Department of Health Outcomes Research and Policy, Auburn University Harrison College of Pharmacy (BT, JQ), Auburn, AL; Department of Mathematics and Statistics, Auburn University College of Sciences and Mathematics (YL, JZ), Auburn, AL.

Source of Funding: None.

Prior Presentation: The abstract of this manuscript was accepted for poster presentation at The Professional Society for Health Economics and Outcomes Research (ISPOR 2022) conference, held May 15-18, 2022, in Washington, DC.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (BT, YL, JZ, JQ); acquisition of data (BT, JQ); analysis and interpretation of data (BT, JZ); drafting of the manuscript (BT); critical revision of the manuscript for important intellectual content (BT, YL, JZ, JQ); statistical analysis (BT, YL); provision of patients or study materials (BT, JQ); administrative, technical, or logistic support (BT); and supervision (JQ).

Address Correspondence to: Jingjing Qian, PhD, Department of Health Outcomes Research and Policy, Auburn University Harrison College of Pharmacy, 4306d Walker Building, Auburn, AL 36849. Email: jzq0004@auburn.edu.

REFERENCES

1. CDC. National Diabetes Statistics Report, 2020: estimates of diabetes and its burden in the United States. 2020. Accessed April 1, 2022. https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics-report.pdf

2. CDC. National Diabetes Statistics Report, 2020: incidence of newly diagnosed diabetes. 2020. Accessed April 1, 2022. https://www.cdc.gov/diabetes/data/statistics-report/newly-diagnosed-diabetes.html

3. American Diabetes Association. Economic costs of diabetes in the U.S. in 2017. Diabetes Care. 2018;41(5):917-928. doi:10.2337/dci18-0007

4. Diabetes: key facts. World Health Organization. Updated April 5, 2023. Accessed December 6, 2021. https://www.who.int/news-room/fact-sheets/detail/diabetes

5. Inzucchi SE, Bergenstal RM, Buse JB, et al; American Diabetes Association; European Association for the Study of Diabetes. Management of hyperglycemia in type 2 diabetes: a patient-centered approach; position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care. 2012;35(6):1364-1379. doi:10.2337/dc12-0413

6. American Diabetes Association. 9. Pharmacologic approaches to glycemic treatment: Standards of Medical Care in Diabetes-2021. Diabetes Care. 2021;44(suppl 1):S111-S124. doi:10.2337/dc21-S009

7. DeFronzo RA, Ratner RE, Han J, Kim DD, Fineman MS, Baron AD. Effects of exenatide (exendin-4) on glycemic control and weight over 30 weeks in metformin-treated patients with type 2 diabetes. Diabetes Care. 2005;28(5):1092-1100. doi:10.2337/diacare.28.5.1092

8. Dicker D. DPP-4 inhibitors: impact on glycemic control and cardiovascular risk factors. Diabetes Care. 2011;34(suppl 2):S276-S278. doi:10.2337/dc11-s229

9. Fala L. Invokana (canagliflozin): first-in-class SGLT2 inhibitor approved for the treatment of type 2 diabetes. 2013. Accessed April 1, 2022. https://www.ahdbonline.com/web-exclusives/fda-approvals/1575-invokana-canagliflozin-first-in-class-sglt2-inhibitor-approved-for-the-treatment-of-type-2-diabetes

10. Empagliflozin (Jardiance). GoodRx. Accessed December 6, 2021. https://www.goodrx.com/empagliflozin

11. Exenatide (Byetta). GoodRx. Accessed December 6, 2021. https://www.goodrx.com/byetta

12. Sitagliptin (Januvia). GoodRx. Accessed December 6, 2021. https://www.goodrx.com/sitagliptin

13. DeJong C, Masuda C, Chen R, Kazi DS, Dudley RA, Tseng CW. Out-of-pocket costs for novel guideline-directed diabetes therapies under Medicare Part D. JAMA Intern Med. 2020;180(12):1696-1699. doi:10.1001/jamainternmed.2020.2922

14. Liss DT, Kang RH, Lancki N, et al. Costs for commercially insured adults prescribed second-line diabetes medications. Am J Manag Care. 2021;27(3):e72-e79. doi:10.37765/ajmc.2021.88601

15. Riddle MC, Herman WH. The cost of diabetes care-an elephant in the room. Diabetes Care. 2018;41(5):929-932. doi:10.2337/dci18-0012

16. Gilstrap LG, Blair RA, Huskamp HA, Zelevinsky K, Normand SL. Assessment of second-generation diabetes medication initiation among Medicare enrollees from 2007 to 2015. JAMA Netw Open. 2020;3(5):e205411. doi:10.1001/jamanetworkopen.2020.5411

17. Le P, Chaitoff A, Misra-Hebert AD, Ye W, Herman WH, Rothberg MB. Use of antihyperglycemic medications in U.S. adults: an analysis of the national health and nutrition examination survey. Diabetes Care. 2020;43(6):1227-1233. doi:10.2337/dc19-2424

18. Fang M, Wang D, Coresh J, Selvin E. Trends in diabetes treatment and control in U.S. adults, 1999-2018. N Engl J Med. 2021;384(23):2219-2228. doi:10.1056/NEJMsa2032271

19. Johnson CL, Paulose-Ram R, Ogden CL, et al. National Health and Nutrition Examination Survey: analytic guidelines, 1999-2010. Vital Health Stat 2. 2013;(161):1-24.

20. American Diabetes Association. Classification and diagnosis of diabetes: Standards of Medical Care in Diabetes-2021. Diabetes Care. 2021;44(suppl 1):S15-S33. doi:10.2337/dc21-S002

21. Demmer RT, Zuk AM, Rosenbaum M, Desvarieux M. Prevalence of diagnosed and undiagnosed type 2 diabetes mellitus among US adolescents: results from the continuous NHANES, 1999-2010. Am J Epidemiol. 2013;178(7):1106-1113. doi:10.1093/aje/kwt088

22. National Health and Nutrition Examination Survey: MEC laboratory procedures manual. CDC. January 2017. Accessed April 1, 2022. https://wwwn.cdc.gov/nchs/data/nhanes/2017-2018/manuals/2017_MEC_Laboratory_Procedures_Manual.pdf

23. Dietary Guidelines for Americans, 2020-2025. 9th ed. US Department of Agriculture; December 2020. Accessed April 1, 2022. https://www.dietaryguidelines.gov/resources/2020-2025-dietary-guidelines-online-materials

24. National Health and Nutrition Examination Survey: 2011-2012 data documentation, codebook, and frequencies: physical activity (PAQ_G). CDC. September 2013. Accessed January 15, 2020. https://wwwn.cdc.gov/Nchs/Nhanes/2011-2012/PAQ_G.htm

25. About adult BMI. CDC. Updated June 3, 2022. Accessed January 15, 2021. https://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/index.html

26. Stevens PE, Levin A; Kidney Disease: Improving Global Outcomes Chronic Kidney Disease Guideline Development Work Group Members. Evaluation and management of chronic kidney disease: synopsis of the Kidney Disease: Improving Global Outcomes 2012 clinical practice guideline. Ann Intern Med. 2013;158(11):825-830. doi:10.7326/0003-4819-158-11-201306040-00007

27. Levey AS, Stevens LA, Schmid CH, et al; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. doi:10.7326/0003-4819-150-9-200905050-00006

28. Selvin E, Manzi J, Stevens LA, et al. Calibration of serum creatinine in the National Health and Nutrition Examination Surveys (NHANES) 1988-1994, 1999-2004. Am J Kidney Dis. 2007;50(6):918-926. doi:10.1053/j.ajkd.2007.08.020

29. National Health and Nutrition Examination Survey: sample design. CDC. Accessed December 13, 2021. https://wwwn.cdc.gov/nchs/nhanes/tutorials/sampledesign.aspx

30. National Health and Nutrition Examination Survey: weighting. CDC. Accessed December 13, 2021. https://wwwn.cdc.gov/nchs/nhanes/tutorials/weighting.aspx

31. Chang TI, Park H, Kim DW, et al. Polypharmacy, hospitalization, and mortality risk: a nationwide cohort study. Sci Rep. 2020;10(1):18964. doi:10.1038/s41598-020-75888-8

32. Corsonello A, Pedone C, Corica F, Incalzi RA. Polypharmacy in elderly patients at discharge from the acute care hospital. Ther Clin Risk Manag. 2007;3(1):197-203. doi:10.2147/tcrm.2007.3.1.197

33. McIsaac DI, Wong CA, Bryson GL, van Walraven C. Association of polypharmacy with survival, complications, and healthcare resource use after elective noncardiac surgery: a population-based cohort study. Anesthesiology. 2018;128(6):1140-1150. doi:10.1097/ALN.0000000000002124

34. de Decker L, Hanon O, Boureau AS, et al. Association between hypoglycemia and the burden of comorbidities in hospitalized vulnerable older diabetic patients: a cross-sectional, population-based study. Diabetes Ther. 2017;8(6):1405-1413. doi:10.1007/s13300-017-0319-7

35. Lee SE, Kim KA, Son KJ, et al. Trends and risk factors in severe hypoglycemia among individuals with type 2 diabetes in Korea. Diabetes Res Clin Pract. 2021;178:108946. doi:10.1016/j.diabres.2021.108946

36. Han K, Yun JS, Park YM, et al. Development and validation of a risk prediction model for severe hypoglycemia in adult patients with type 2 diabetes: a nationwide population-based cohort study. Clin Epidemiol. 2018;10:1545-1559. doi:10.2147/CLEP.S169835

37. Foley JE, Jordan J. Weight neutrality with the DPP-4 inhibitor, vildagliptin: mechanistic basis and clinical experience. Vasc Health Risk Manag. 2010;6:541-548. doi:10.2147/vhrm.s10952

38. Novo Nordisk receives FDA approval for Saxenda (liraglutide [rDNA origin] injection) for chronic weight management. News release. PR Newswire. December 24, 2014. Accessed April 1, 2022. https://www.prnewswire.com/news-releases/novo-nordisk-receives-fda-approval-for-saxenda-liraglutide-rdna-origin-injection-for-chronic-weight-management-300013975.html

39. Buse JB, Drucker DJ, Taylor KL, et al; DURATION-1 Study Group. DURATION-1: exenatide once weekly produces sustained glycemic control and weight loss over 52 weeks. Diabetes Care. 2010;33(6):1255-1261. doi:10.2337/dc09-1914

40. FDA approves new drug treatment for chronic weight management, first since 2014. News release. FDA. June 4, 2021. Accessed April 1, 2022. https://www.fda.gov/news-events/press-announcements/fda-approves-new-drug-treatment-chronic-weight-management-first-2014

41. Bays HE, Weinstein R, Law G, Canovatchel W. Canagliflozin: effects in overweight and obese subjects without diabetes mellitus. Obesity (Silver Spring). 2014;22(4):1042-1049. doi:10.1002/oby.20663