The epidemic of type 2 diabetes (T2D) in the United States incurs significant human and economic costs, including severe long-term complications such as retinopathy, nephropathy, neuropathy, and cardiovascular disease (CVD). Although research has led to the creation of new classes of glucose-lowering medications, selecting the most effective way to achieve proper glycemic control among the variety of treatment options presents a considerable challenge for healthcare providers. The Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE) was a practical, unmasked clinical trial that evaluated the effectiveness of commonly prescribed diabetes medications—when used alongside metformin—on lowering glycemia and enhancing patient-centered outcomes.

Individuals with duration of T2D of less than 10 years and an age of at least 30 years at time of diagnosis were eligible for the study. Participants completed a run-in phase to adjust metformin to a dose of at least 1,000 mg/day and a target dose of 2,000 mg/day. Following the run-in period, participants with baseline hemoglobin A1c values of 6.8-8.5% (51-69 mmol/mol) were randomized to treatment with metformin and one of four medications: sulfonylureas, dipeptidyl peptidase 4 inhibitors, glucagon-like receptor agonists, and insulin. All participants continued follow-up for a period of 4-7 years (depending on the time of entry), including those who reached the primary outcome.

Continuous Glucose Monitoring (CGM) data are not included in the package, but are available upon request.

The primary objective of the study was comparison of the relative effectiveness of four commonly used glucose-lowering medications with different mechanisms of action, when added to metformin, in maintaining metabolic control, defined as time-to-primary failure with an HbA1c ≥ 7.0%, confirmed, while on maximally tolerated doses of both metformin, up to 2,000 mg per day, and the assigned medication.

The secondary objectives assessed time to secondary and tertiary metabolic failure, treatment durability of the glycemia-lowering effects and other metabolic outcomes, adverse effects, cumulative incidence of diabetic complications, effects on CVD risk factors and quality-of-life, beta-cell function over time, tolerability, and cost-effectiveness.

The primary outcome measure was the time to primary metabolic failure of the randomly assigned treatment, defined as the time to an initial HbA1c ≥ 7%, subsequently confirmed at the next quarterly visit. Secondary outcome measures included mean HbA1c and fasting glucose levels, frequency of hypoglycemia, cardiovascular events, estimated glomerular filtration rate (eGFR), albuminuria, adverse events, and mortality.

Inclusion Criteria

• Participants diagnosed with diabetes at age ≥ 30 years ( ≥ 20 for American Indians)
• Duration of diagnosed diabetes < 10 years
• HbA1c criteria (at final run-in visit, ~2 weeks prior to randomization): 6.8-8.5%
• Taking a daily dose of ≥ 1,000 mg metformin for a minimum of 8 weeks at final run-in
• Willingness to administer daily subcutaneous injections, take a second diabetes drug after randomization, potentially initiate insulin and intensify insulin therapy if study metabolic goals are not met, perform self-monitoring of blood glucose
• Fluent in either English or Spanish
• A negative pregnancy test for all females of childbearing potential (i.e., pre-menopausal, and not surgically sterile)
• Provision of signed and dated informed consent prior to any study procedures

Exclusion Criteria

• Suspected type 1 diabetes (lean with polyuria, polydipsia, and weight loss with little response to metformin) or “secondary” diabetes due to specific causes (e.g., previously diagnosed monogenic syndromes, pancreatic surgery, pancreatitis)
• Current or previous (within past 6 months) treatment with any diabetes drug/glucose-lowering medication other than metformin (limited use of no longer than seven days was allowed, for example during hospitalization)
• More than 10 years of treatment with metformin at time of randomization screening
• History of intolerance or allergy or other contraindications to any of the proposed study medications
• Participation in another interventional clinical trial, previous randomization in the GRADE study, or resides in the same household with another GRADE study participant
• Current need for any specific glucose-lowering medications solely for other conditions, for example for polycystic ovary syndrome
• Symptomatic hyperglycemia requiring immediate therapy during screening or run-in, in the judgment of the physician
• A life-threatening event within 30 days prior to screening or currently planned major surgery
• Any major cardiovascular event in previous year, including history of myocardial infarction, stroke, or vascular procedure such as coronary artery or peripheral bypass grafting, stent placements (peripheral or coronary), or angioplasty
• Plans for pregnancy during the study for women of child-bearing potential
• History of or planning bariatric surgery, including banding procedures or surgical gastric and/or intestinal bypass (if banding removed, may be considered eligible after 1 year)
• History of congestive heart failure (NYHA 3 or greater); pancreatitis; severe liver disease or acute hepatitis or ALT > 3 times upper limit of normal; hemolytic anemia, chronic transfusion requirement, or other condition rendering HbA1c results unreliable as indicator of chronic glucose levels; or cancer, other than non-melanoma skin cancer, that required therapy in the 5 years prior to randomization
• Personal or family history of MEN-2 or family history of medullary thyroid cancer
• Estimated GFR (eGFR) < 30 ml/min/1.73 m2 or end stage renal disease requiring renal replacement therapy
• Current alcoholism or excessive alcohol intake
• Previous organ transplant
• Treatment with atypical antipsychotics, or with oral or systemic glucocorticoids (other than short-term treatment, for example for poison ivy) or disease likely to require periodic or regular glucocorticoid therapy (inhaled steroids were allowed)
• Clinically or medically unstable with expected survival < 1 year
• No non-study PCP or inability to identify such a PCP (who will provide non-study care) by the time of final run-in, or unwillingness to permit sites to contact PCP
• In the opinion of the principal investigator (PI), any other factor, including language barrier, likely to limit compliance with the protocol

The GRADE study enrolled and followed 5,047 participants with type 2 diabetes < 10 years duration (19.8% Black, 18.6% Hispanic) taking metformin for a mean of 5.0 years at 36 participating clinical centers in the United States. Participants were randomized to one of four glucose lowering medications, insulin glargine U-100, the sulfonylurea glimepiride, the glucagon-like peptide-1 receptor agonist liraglutide, or the dipeptidyl peptidase 4 inhibitor sitagliptin. All participants were followed with quarterly visits and assessed for the primary and secondary metabolic outcomes, HbA1c ≥ 7.0%, HbA1c > 7.5, respectively, based on centrally measured glycated hemoglobin and confirmed at the next visit. The primary outcome was significantly different (p < 0.001) among the treatment groups over the mean 5 years of follow-up. The rates for the glargine and liraglutide groups were similar and lower compared to the other groups, glimepiride and sitagliptin. The analysis of the secondary metabolic showed similar results. The reports of side effects varied among treatment groups. Risk of severe hypoglycemia was rare with rates ranging from 0.7% for participants assigned to sitagliptin to 2.2% for glimepiride participants. For glycemia outcomes, participants in all treatment groups had an initially decreased HbA1c followed by a gradual increase towards metabolic failure, with glargine and liraglutide being slightly more effective at prolonging the period before metabolic failure over the study follow-up.

The GRADE study also examined the risk of microvascular and cardiovascular disease outcomes and found there were no differences among treatment groups for microvascular outcomes (hypertension, dyslipidemia, albuminuria, renal impairment, or diabetes peripheral neuropathy). There were no treatment group differences for major adverse cardiovascular events (MACE), hospitalization for heart failure, death from cardiovascular causes, or all-cause mortality. There were small differences in the hazard ratios (HR) for any cardiovascular disease when either glargine, glimepiride, or sitagliptin was compared to the other groups combined, whereas the HR was significantly lower in the liraglutide group, 0.7 (95% CI, 0.6-0.9) compared to the other groups combined.