Summary: In an open-label, single-centre randomised trial of 100 adults with type 2 diabetes inadequately controlled on triple oral therapy (metformin, a sulfonylurea, and a DPP-4 inhibitor), adding insulin glargine produced a significantly greater reduction in glycaemic variability than adding dapagliflozin 10 mg over 3 months. Both arms improved across continuous-glucose-monitoring and standard glycaemic measures; the abstract reports no effect sizes, confidence intervals, or p-values.
PICO Summary
| Element | Detail |
|---|---|
| Population | 100 adults with type 2 diabetes inadequately controlled on triple drug therapy (metformin, sulfonylurea, DPP-4 inhibitor); single-centre, India; 1-year prospective, randomised, open-label, double-arm interventional study. |
| Intervention | Insulin glargine added as the fourth glucose-lowering agent for 3 months (n=50). |
| Comparison | Dapagliflozin 10 mg once daily added as the fourth glucose-lowering agent for 3 months (n=50). |
| Outcome | Glycaemic variability assessed by flash continuous glucose monitoring (coefficient of variation, time in range, time above range, time below range, glucose management indicator). Both arms improved variability and standard glycaemic parameters (fasting and postprandial glucose, HbA1c, body weight) at 3 months. Insulin glargine achieved a significantly greater reduction in glycaemic variability than dapagliflozin. The abstract reports no numerical effect sizes, 95% confidence intervals, p-values, or arm-level hypoglycaemia or weight figures. |
Expert Commentary
This small single-centre trial offers a hypothesis-generating signal rather than a practice-changing result. The verdict is that basal insulin appears to flatten glycaemic variability more than dapagliflozin when either is bolted onto failing triple oral therapy, which is mechanistically unsurprising given that titrated basal insulin directly addresses fasting hyperglycaemia. The central limitation is methodological fragility. With only 100 participants randomised, an open-label design, a single centre, and a 3-month window, the study is vulnerable to performance and detection bias, and the absence of reported effect sizes, confidence intervals, or p-values in the abstract means the magnitude and precision of the difference cannot be judged. No arm-level hypoglycaemia or weight data are quantified, so the older framing that insulin caused more hypoglycaemia while dapagliflozin helped weight is not supported by what is reported here. Can I use this with my patients? Not yet as a reason to favour insulin over an SGLT2 inhibitor, because dapagliflozin carries cardiorenal benefits that this short variability-focused trial was not designed to capture, and those outcomes usually drive fourth-agent selection. Larger, blinded, multicentre trials reporting full statistics and patient-centred endpoints are needed before glycaemic-variability data should reshape sequencing decisions.
References
Bhosle DS, Chandekar B, Shelke S. Comparative study of efficacy and safety of dapagliflozin vs basal insulin in stabilizing glycemic variability in patients with type 2 diabetes mellitus inadequately controlled on triple drug combination therapy. J Assoc Physicians India. 2025;73(4):29-35. doi:10.59556/japi.73.0925
