Summary:
In adults with type 2 diabetes of less than 10 years’ duration inadequately controlled on metformin monotherapy, the addition of insulin glargine, glimepiride, liraglutide, or sitagliptin significantly maintained high medication adherence and improved glycemic outcomes across all groups compared to each other, with minor differences observed at year 3 favoring glimepiride and sitagliptin over liraglutide, though it was associated with clinically insignificant variations in adherence between groups.
| PICO | Description |
|---|---|
| Population | Adults with type 2 diabetes (duration <10 years), HbA1c 6.8%-8.5%, treated with metformin monotherapy. |
| Intervention | Addition of second-line agents: insulin glargine, glimepiride, liraglutide, or sitagliptin; adherence measured over 3 years using a validated scale. |
| Comparison | Differences in medication adherence and glycemic outcomes between the four treatment arms. |
| Outcome | Overall adherence was high (mean 88.7/100) across all treatment groups. A 10-point drop in adherence was associated with a 15–19% increased risk of poor glycemic control (HbA1c ≥7.0% or >7.5%). Glimepiride and sitagliptin showed slightly better adherence than liraglutide at 3 years (P<0.05), but differences were small and not clinically significant. |
Clinical Context
Medication adherence represents a critical yet modifiable factor in diabetes management, with non-adherence contributing to poor glycemic control and increased complications. The GRADE study, one of the largest comparative effectiveness trials in type 2 diabetes, evaluated four commonly used second-line agents added to metformin. While efficacy data from clinical trials guides medication selection, real-world adherence patterns significantly influence treatment success. Different medication classes impose varying burdens on patients—from daily injections to multiple daily pills—potentially affecting long-term adherence. Understanding adherence patterns across treatment options helps clinicians make evidence-based decisions that account for both efficacy and real-world patient behavior. This analysis from GRADE provides robust, head-to-head adherence data over three years, filling a critical gap in diabetes therapeutics research.
Clinical Pearls
- High Overall Adherence: All four medication classes (insulin glargine, glimepiride, liraglutide, sitagliptin) achieved remarkably high adherence (mean 88.7/100) over 3 years, challenging assumptions about poor real-world adherence.
- Adherence-Outcome Link: Each 10-point drop in adherence increased risk of inadequate glycemic control by 15-19%, quantifying the clinical importance of adherence support.
- Injectable vs Oral: Despite being injectable, liraglutide showed only marginally lower adherence than oral glimepiride and sitagliptin at year 3, with differences too small to guide treatment selection.
- Clinical Significance: Small statistical differences in adherence between groups lacked clinical significance, suggesting medication selection should prioritize efficacy, safety, and patient preferences over adherence concerns.
Practical Application
When selecting second-line agents for type 2 diabetes, clinicians can reassure patients that adherence rates are similarly high across insulin, sulfonylureas, GLP-1 agonists, and DPP-4 inhibitors when used in motivated trial participants. The strong association between adherence and glycemic control underscores the value of adherence assessment and support in clinical practice. Use validated adherence measures at follow-up visits, and intervene when scores drop. For patients expressing concerns about injection burden, emphasize that liraglutide adherence was comparable to oral agents. Focus discussions on medication-specific benefits (cardiovascular protection, weight effects, hypoglycemia risk) rather than presumed adherence differences. Recognize that real-world adherence may be lower than trial populations; implement systems-based approaches like pharmacy refill monitoring, medication synchronization, and regular adherence counseling to maintain the high adherence rates demonstrated in GRADE.
Broader Evidence Context
These GRADE findings contradict common assumptions that injectable medications have inferior adherence compared to oral agents. Previous meta-analyses have shown wide variability in diabetes medication adherence (30-80%), but most relied on retrospective pharmacy data or observational studies with heterogeneous populations. GRADE’s prospective, standardized adherence measurement across multiple drug classes provides higher-quality comparative data. The finding that adherence predicts glycemic control aligns with extensive prior literature but quantifies this relationship specifically for second-line agents in a controlled setting. Interestingly, the minimal differences between medication classes suggest patient factors (motivation, diabetes education, healthcare access) may outweigh medication-specific factors in determining adherence patterns.
Study Limitations
- Trial participants received intensive support and education, potentially inflating adherence rates beyond real-world practice.
- Self-reported adherence measure may overestimate actual medication-taking behavior compared to objective measures like pharmacy refills or electronic monitoring.
- Short diabetes duration (<10 years) and relatively controlled baseline HbA1c may limit generalizability to more advanced or poorly controlled patients.
- Analysis doesn’t account for medication-specific side effects that might affect adherence over longer periods.
- Cost and insurance coverage, major real-world adherence barriers, were controlled in trial setting.
Bottom Line
In adults with type 2 diabetes inadequately controlled on metformin, adding insulin glargine, glimepiride, liraglutide, or sitagliptin results in similarly high medication adherence over three years. Better adherence strongly predicts improved glycemic control across all medication classes. Clinicians should select second-line agents based on patient-specific efficacy and safety considerations rather than concerns about differential adherence between medication classes.
Source: Gonzalez, Jeffrey S., et al. “Medication Adherence in the Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study (GRADE).” Diabetes Care, 2026 Feb 1;49(2):335-343. Read article here.
