Mexico is one of the three countries in America with the highest prevalence (14.4%) of DM. Providing medical care to uncontrolled T2DM patients who failed antidiabetic treatment (ADT) is challenging. We designed a retrospective study to assess the response to dapagliflozin (Dapa) or DPP-4-inhibitors (DPP-4i) as add-on therapy in uncontrolled T2DM patients who had previous ADT.

Objectives: Compare subjects with ≥5% improvement of the weighted sum of the composite variable (HbA1c, body weight (BW) and blood pressure (BP)) by the end of the 24-week observation period.

Inclusion Criteria: T2DM subjects ≥18 years old with failure to any ADT and HbA1c% ≥7 that started Dapa or DPP-4i as add-on to other antidiabetic drugs from August 2014 to November 2017.

Statistical Methods and Analysis: A propensity score matching model (nearest neighbor algorithm) was used, the covariates included were prior ADT, TC, LDL-c, HbA1c, BW, SBP and age.

Results: Of 1379 subjects, 131 met statistical matching criteria. Subjects with Dapa had a significant improvement (>5%) in the composite variable compared to DPP-4i (60.3% vs. 34.3%, p<0.01). Significant results were shown in BMI (-0.36 (-1.04 to 0.40) vs. 0 (-0.62 to 0.78), p<0.01), weight (-1 (-3 to 1) vs. 0 (-1.69 a 2), p<0.001), glucose (-8 (-39 to 24) vs. -1 (-22 to 16), p<0.01) and HbA1c (-0.89 (-1.44 to -0.38) vs. -0.12 (-0.82 to 0.46), p<0.01). Logistic Regression Model (LRM) was carried out to identify if the improvement at Dapa group was independent to the confusion factors. LRM confirmed that the improvement in the primary outcome was dependent of the type therapy, regardless of age, BMI, triglyceride, HDL and LDL cholesterol levels and ADT at the beginning of the study (Dapa OR 4.23 IC 95% 2.46-7.29, p<0.001).

Conclusion: T2DM Mexican patients who have failed to previous antidiabetic therapies showed significant decrements of HbA1c, glucose, weight and blood pressure levels with the use of Dapa.


D. Espinoza-Peralta: None. A.I. Silva: Employee; Self; AstraZeneca. J.G. Collado: None. J.D. Rogel-Manzanares: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at