The effect of real-time continuous glucose monitoring on glycemic control in patients with type 2 diabetes mellitus

Abstract

Background: Real-time continuous glucose monitoring (RT-CGM) improves hemoglobin A1c (A1C) and hypoglycemia in people with type 1 diabetes mellitus and those with type 2 diabetes mellitus (T2DM) on prandial insulin; however, it has not been tested in people with T2DM not taking prandial insulin. We evaluated the utility of RT-CGM in people with T2DM on a variety of treatment modalities except prandial insulin.

Methods: We conducted a prospective, 52-week, two-arm, randomized trial comparing RT-CGM (n = 50) versus self-monitoring of blood glucose (SMBG) (n = 50) in people with T2DM not taking prandial insulin. Real-time continuous glucose monitoring was used for four 2-week cycles (2 weeks on/1 week off). All patients were managed by their usual provider. This article reports on changes in A1C 0-12 weeks.

Results: Mean (± standard deviation) decline in A1C at 12 weeks was 1.0% (± 1.1%) in the RT-CGM group and 0.5% (± 0.8%) in the SMBG group (p = .006). There were no group differences in the net change in number or dosage of hypoglycemic medications. Those who used the RT-CGM for ≥ 48 days (per protocol) reduced their A1C by 1.2% (± 1.1%) versus 0.6% (± 1.1%) in those who used it <48 days (p = .003). Multiple regression analyses statistically adjusting for baseline A1C, an indicator for usage, and known confounders confirmed the observed differences between treatment groups were robust (p = .009). There was no improvement in weight or blood pressure.

Conclusions: Real-time continuous glucose monitoring significantly improves A1C compared with SMBG in patients with T2DM not taking prandial insulin. This technology might benefit a wider population of people with diabetes than previously thought.

© 2011 Diabetes Technology Society.

Figures

Figure 1

Study design.

Figure 2

Changes in A1C from baseline to 12 weeks, (A) by treatment group and (B) by RT-CGM usage group. The figure shows boxplots. The boxes themselves contain the 25th, 50th, and 75th percentiles. The whiskers of the boxes show the minimum and maximum values, with the dots beneath the whiskers indicating possible outlying values. Means ± SDs are shown in parentheses within each box. P values are from t-tests or ANOVA comparing the group's mean changes. In multiple regression analyses in which change scores for each outcome were regressed on age, gender, therapy, baseline value for the outcome, and treatment group, group was a significant predictor of change in A1C, -0.48 meaning the RT-CGM group had an adjusted decline in A1C of 0.48% greater than the SMBG group (p = .006). For the RT-CGM group ≥ 48 days, adjusted decline in A1C was -0.60 (p = .002) relative to the SMBG group. Group was not significant for the other outcomes, weight, and BP.

Figure 3

Interaction of baseline A1C with group in the prediction of 12-week change in A1C. Figures were derived from multiple regression analyses. P values for interaction effects are shown in parentheses in the graphs and refer to the difference from the SMBG group. The coefficient (standard error) for the A interaction term, group × baseline A1C, was -0.39 (0.14). The coefficient (standard error) for the B interaction terms, (RT-CGM < 48 days) × (A1C at baseline) and (RT-CGM ≥ 48 days) × (A1C at baseline), were -0.36 (0.23) and -0.40 (0.15), respectively. Equations for the plotting of the lines assumed mean age, male gender, and oral medications only.