Factors Associated with HbA1c-Lowering in a Randomized Trial of Continuous Glucose Monitoring in Adults with Non-Insulin-Treated Diabetes: An Observational Analysis.

Background

A prior randomized trial showed that CGM lowers HbA1c in adults with type 2 diabetes and elevated glycated hemoglobin (HbA1c > 7.0%) not on insulin.

Objectives

We examined the effect modifiers of CGM effectiveness, and factors associated with HbA1c-lowering.

Methods

Secondary analysis of data from a randomized trial. The primary outcome was change in HbA1c from baseline to 12 weeks. Explanatory variables included sociodemographics, diabetes and treatment parameters, and neighbourhood environment characteristics. Variables were assessed for interaction with CGM using linear regression, which was also used to estimate associations between each variable and HbA1c change, and to develop a final multi-variable adjusted-model.

Results

We included 86 individuals. No formal interactions were identified with CGM effectiveness, though noticeable numeric differences in the effect of CGM were seen with sex, recent diagnosis, and neighbourhood walkability. Aside from CGM treatment allocation (β=-0.65, p=0.003 95% CI [-1.07 to -0.22]), multi-variable predictors of HbA1c-lowering were higher baseline HbA1c (β =-0.71, p<0.001 [-0.-90 to -0.53]), diabetes diagnosis < 1 year (β=-1.06, p=0.004 [-1.77 to -0.35]), male sex (β=-0.51, p=0.017 [-0.93 to -0.09]), and not using sodium-glucose co-transporter-2 inhibitor medication at baseline (β=0.62, p=0.006 [0.18 to 1.05]). Neighbourhood walkability was associated with HbA1c-lowering on crude (β=-0.26, p=0.041 [-0.51 to -0.01]), but not adjusted, analysis.

Conclusions

Further studies are needed to determine the factors influencing the effectiveness of CGM. Individuals with type 2 diabetes, not on insulin, who voluntarily seek CGM, may experience a reduction in HbA1c proportional to their baseline HbA1c, regardless of treatment allocation.