Recent evidence shows that many patients with type 2 diabetes (T2DM) initially decline insulin treatment. For some patients this could lead to poor glycemic control, whereas others may be able to lower blood glucose by alternative means. However, it is not currently known what factors affect achievement of glycemic control after decline of insulin therapy by patients.

We conducted a retrospective cohort study to determine patient characteristics predictive of glycemic control after they decline insulin treatment. We studied adults with T2DM treated at two academic medical centers between 1993 and 2014 who rejected insulin therapy recommended by their providers. We compared 150 randomly selected patients whose A1c subsequently decreased to 150 patients whose A1c increased or stayed the same.

Patients whose A1c decreased had a significantly higher baseline A1c (9.5% vs. 8.5%; p < .001), were more likely to implement lifestyle changes (27.3% vs. 7.3%; p < .001) and less likely to be non-adherent to diabetes medications (1.3% vs. 26.7%; p < .001) or to discontinue a non-insulin diabetes medication (8.0% vs. 16.0%; p = 0.03) after they declined insulin therapy.

In a multivariable analysis adjusted for demographics, comorbidities and clustering within providers, higher baseline A1c (OR 1.85; 95% CI 1.41 to 2.42; p <.001) and lifestyle changes (OR 8.49; 95% CI 3.31 to 21.78; p <0.001) were associated with greater, while non-adherence to diabetes medications (OR 0.015; 95% CI 0.0025 to 0.088; p <.001) and discontinuation of a non-insulin diabetes medication (OR 0.30; 95% CI 0.12 to 0.78; p =0.013) were associated with lower probability of A1c decrease after initial decline of insulin therapy by the patient.

We identified patient characteristics and treatment strategies associated with success and failure of glycemic control after insulin therapy decline by the patient. This information can help guide care and assist in design of rigorous interventional studies to evaluate optimal therapeutic approaches for these individuals.


L. Florez: None. A. Turchin: Advisory Panel; Self; Monarch Medical Technologies. Research Support; Self; Eli Lilly and Company. Stock/Shareholder; Self; Brio Systems.

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