…

The DIVERSITY study arrives at a moment when clinicians increasingly recognize that the populations enrolled in randomized controlled trials rarely resemble the patients they see every day. Trial participants are typically younger, healthier, and more adherent than the average person with type 2 diabetes, and they are carefully screened to exclude the very comorbidities—chronic kidney disease, hepatic impairment, advanced age, polypharmacy—that complicate real-world prescribing decisions. Non-interventional studies like DIVERSITY cannot replace randomized evidence for establishing causality, but they answer a different and equally important question: what happens when a therapy is deployed across the messy, heterogeneous spectrum of routine clinical practice, where physicians treat patients with varying risk profiles, treatment histories, and levels of glycemic control?

The geographic scope of the study deserves attention. By enrolling patients across seven countries in Central, Eastern, and Southeastern Europe—Bosnia and Herzegovina, Bulgaria, Croatia, Hungary, North Macedonia, Romania, and Slovenia—the investigators captured a region where diabetes prevalence has been rising steadily and where healthcare systems differ substantially in resources, reimbursement structures, and specialist availability. Much of the published real-world evidence on dipeptidyl peptidase-4 inhibitors has come from Western Europe, North America, or East Asia, so data from these healthcare environments add a genuinely underrepresented perspective to the evidence base. The involvement of 206 investigators, including diabetologists, endocrinologists, and other specialists, further reflects the reality that type 2 diabetes is managed not only in dedicated diabetes clinics but also in general internal medicine and primary care settings.

The mechanism of action underlying sitagliptin’s effects helps explain both its efficacy and its favorable tolerability profile. Dipeptidyl peptidase-4 inhibitors work by prolonging the activity of incretin hormones—glucagon-like peptide-1 and glucose-dependent insulinotropic polypeptide—that are released by the gut in response to food intake. These hormones stimulate insulin secretion from pancreatic beta cells and suppress glucagon release, but critically, both effects are glucose-dependent. Insulin secretion is stimulated only when blood glucose is elevated, and glucagon suppression occurs only in the hyperglycemic state. This built-in feedback mechanism is the principal reason why DPP-4 inhibitors, as a class, carry a very low intrinsic risk of hypoglycemia when used alone or with agents that do not independently drive insulin release indiscriminately. The near-absence of treatment-related hypoglycemic symptoms in DIVERSITY—just three cases among 2,697 treated patients over six months—is consistent with this pharmacology, though it is worth noting that patients on stable insulin therapy were included in the study, a population in which hypoglycemia risk would ordinarily be dominated by the insulin itself.

The distinction between fasting and post-prandial glucose improvements in the study is also mechanistically meaningful. Sitagliptin reduced fasting plasma glucose by 1.91 mmol/L and post-prandial glucose by 2.39 mmol/L over the six-month observation period. The relatively larger effect on post-prandial values aligns with the incretin pathway’s role, since incretin hormones are released primarily in response to meals and their activity is most pronounced in the post-prandial state. Post-prandial hyperglycemia has gained increasing attention in diabetes research as an independent contributor to overall glycemic variability and, potentially, to cardiovascular risk, making therapies that address the meal-related glucose excursions particularly relevant even when their HbA1c-lowering magnitude is moderate.

Weight neutrality is another characteristic that distinguishes DPP-4 inhibitors from several other glucose-lowering classes. In DIVERSITY, mean BMI changed by only 0.61 kg/m² over six months, a small shift consistent with the weight-neutral profile established in randomized trials. This matters clinically because weight gain is a common barrier to treatment intensification: sulfonylureas and insulin typically promote weight gain, which can worsen insulin resistance and patient willingness to escalate therapy. Thiazolidinediones carry similar concerns along with fluid retention. By contrast, agents that neither add nor subtract meaningful weight allow clinicians to intensify glycemic control without creating a countervailing metabolic burden. For the many patients in DIVERSITY who were older or had multiple comorbidities, avoiding additional weight gain and hypoglycemia is often as important as the absolute HbA1c reduction itself.

The single-pill combination of sitagliptin with metformin addresses a different but related practical challenge: pill burden and treatment complexity. Metformin remains the foundational first-line pharmacotherapy for type 2 diabetes in most guidelines, and the stepwise intensification approach endorsed by current consensus reports typically involves adding agents to metformin rather than replacing it. When intensification means adding a second or third oral agent, the number of daily tablets can climb quickly, and poor adherence to multi-tablet regimens is a well-documented driver of suboptimal glycemic outcomes. Combining two agents with complementary mechanisms—metformin acting primarily on hepatic glucose production and insulin sensitivity, sitagliptin acting on the incretin axis—into one tablet taken once or twice daily simplifies the regimen without changing the pharmacology. The slightly larger HbA1c reduction observed with the single-pill combination in DIVERSITY (1.03 percentage points versus 0.93 for monotherapy) is consistent with the additive glycemic effects expected when metformin is part of the regimen, though the observational design means the two groups were not randomized and baseline characteristics likely differed.

The proportion of patients reaching the conventional HbA1c target of below 7 percent more than tripled during the study, rising from 16.9 percent at baseline to 53.7 percent at six months. This is a striking shift given that nearly half of patients with type 2 diabetes worldwide fail to achieve recommended glycemic targets despite available therapies. Part of the improvement reflects the simple fact that patients were selected for eligibility to receive these regimens and were followed prospectively with structured data captures, conditions that do not exist in unstructured routine care. Nevertheless, the magnitude and consistency of the shift across both monotherapy and combination regimens suggest that the results are not an artifact of a single favorable subgroup. It is also worth remembering that glycemic targets are individualized in contemporary practice—older patients with limited life expectancy or substantial comorbidity may have targets well above 7 percent—so the true share of patients meeting their own prescribed goals may differ from the share meeting this conventional benchmark.

Treatment acceptability, measured as satisfaction reported by both patients and investigators at six months, exceeded 90 percent in both groups. Patient-reported satisfaction is increasingly recognized as an outcome in its own right, not merely a soft correlate. Therapies that patients find burdensome—because of side effects, dosing complexity, or injection requirements—are associated with poorer adherence and earlier discontinuation, which erodes whatever efficacy the agent possesses under ideal conditions. A therapy that is moderately effective but consistently taken will often outperform a more potent agent that patients abandon. The high satisfaction rates in DIVERSITY, combined with the low symptom burden and simple dosing, fit this logic and help explain why DPP-4 inhibitors have maintained a durable place in treatment algorithms even as newer classes with larger glucose-lowering capacity have entered the market.

The inclusion of older and multimorbid patients is one of the study’s most practically relevant features. Mean age in the analysis set was 64.6 years, and the eligibility criteria deliberately encompassed patients on stable insulin therapy, patients with metformin contraindications or intolerance, and patients requiring intensification beyond metformin plus another oral agent. These are populations frequently excluded from or underrepresented in randomized trials, yet they constitute a large fraction of routine diabetes care. For older patients, the risks of hypoglycemia are amplified—falls, fractures, cognitive impairment, and cardiovascular events can all be triggered or worsened by low glucose episodes—making therapies with low hypoglycemia potential particularly attractive even when their HbA1c-lowering effect is moderate. Similarly, patients with hepatic impairment, for whom sitagliptin has few contraindications, represent a group with limited therapeutic options.

Several limitations inherent to the non-interventional design should temper interpretation of the findings. Without randomization, the choice between sitagliptin monotherapy and the sitagliptin/metformin single-pill combination was made by treating physicians based on clinical judgment, so differences between the regimen groups may reflect differences in the patients themselves rather than differences in the treatments. The six-month follow-up captures early treatment response but cannot speak to durability of glycemic control over years, which is particularly relevant for a progressive disease in which beta-cell function continues to decline. Hypoglycemia was assessed through treatment-related symptoms rather than systematic biochemical monitoring, which may undercount asymptomatic episodes. And the open-label nature of routine practice means that the structured follow-up visits themselves—three data captures over six months—may have introduced a Hawthorne-like effect in which attention and monitoring contribute to improved outcomes independent of pharmacotherapy.

Within these constraints, the DIVERSITY findings situate sitagliptin-based regimens within the contemporary treatment landscape in a useful way. The 2022 ADA/EASD consensus report emphasizes tailoring therapy to cardiovascular and comorbidity profiles, positioning gliptins as add-on options to metformin and sodium–glucose cotransporter 2 inhibitors rather than as agents with organ-protective evidence of their own. In that framework, the value proposition of sitagliptin rests on tolerability, simplicity, weight neutrality, and applicability across broad populations—precisely the attributes that real-world studies like DIVERSITY are designed to test. The results reported here, showing clinically meaningful glycemic improvements with low hypoglycemia incidence and high satisfaction across seven countries and more than 2,600 patients, provide reassurance that the class performs in routine practice much as its trial profile predicted, including in the older and multimorbid patients for whom treatment choices are often most constrained.

Subject of Research: …

Article Title: Real-world outcomes of sitagliptin and sitagliptin/metformin single-pill combination treatment in diverse type 2 diabetes populations (DIVERSITY study)

Article References: Janez, A., Bojic, M., Bozek, T., Kamenov, Z., Milenkovic, T., Radulian, G., & Várkonyi, T. (2026). Real-world outcomes of sitagliptin and sitagliptin/metformin single-pill combination treatment in diverse type 2 diabetes populations (DIVERSITY study). Frontiers in Clinical Diabetes and Healthcare, 7, Article 1905393. https://doi.org/10.3389/fcdhc.2026.1905393

Image Credits: AI Generated

DOI: 10.3389/fcdhc.2026.1905393

Keywords: Real-world, outcomes, sitagliptin, metformin, single-pill, combination, treatment, diverse, type, diabetes, populations, DIVERSITY