Sustained CGM Use Linked to Lower HbA_1c, Urinary Albumin-Creatinine Ratio

Key Highlights

• At 36 months, HbA_1c was approximately 0.8 percentage points lower among patients using real-time continuous glucose monitoring (CGM).
• Urinary albumin-creatinine ratio was 125.3 mg/g lower with sustained monitoring, but estimated glomerular filtration rate did not differ.
• The researchers note “substantial attrition” in patients due to CGM discontinuation in the treatment group and CGM initiation in the control group.

Sustained real-time continuous glucose monitoring (rtCGM) was associated with durable improvements in glycemic control and a later reduction in urinary albumin-creatinine ratio (UACR) among high-risk adults with type 2 diabetes, according to a comparative effectiveness study published in JAMA Network Open. Investigators evaluated outcomes over 36 months to examine longer-term associations with glycemia and kidney markers.

The study emulated a target trial comparing patients who initiated and sustained rtCGM with controls who did not initiate monitoring. Eligible adults were aged 19 years or older, had type 2 diabetes and no prior CGM use, and had either HbA_1c above 8% or emergency department or inpatient treatment for hypoglycemia. Investigators sampled eligible controls monthly from January 2015 through June 2023 and aligned their baseline dates with those of rtCGM initiators.

The researchers used inverse probability-weighted marginal structural models to estimate per-protocol outcomes for HbA_1c, UACR, and estimated glomerular filtration rate (eGFR). Follow-up data were censored when participants deviated from their assigned monitoring strategy, experienced coverage gaps, died, or reached 36 months.

Study Findings

The analysis included 25,648 adults: 2,771 rtCGM initiators and 22,877 control patients. The mean age was 61.5 years, and the mean baseline HbA_1c was 9.5%. At 36 months, HbA1c was 7.83% (95% CI, 7.55%-8.11%) in the rtCGM arm vs 8.58% (95% CI, 8.49%-8.66%) among controls, an approximately 0.8-percentage-point difference (P < .001). Attrition was substantial: 1,228 initiators (44.3%) were censored after discontinuing rtCGM, and 4,284 controls (18.7%) were censored after initiating CGM.

UACR did not differ significantly between groups initially. By 36 months, however, the estimated difference favored rtCGM by 125.3 mg/g (rtCGM minus control, −125.3 mg/g; 95% CI, −196.6 to −54.0; P = .001). No differences in eGFR were observed during follow-up. These results reflect the glycemic and kidney-marker outcomes reported in the study.

Clinical Implications

According to the researchers, sustained rtCGM use was associated with lasting glycemic improvement and later UACR regression. They suggest that addressing glycemic deterioration could delay pharmacologic escalation and potentially slow diabetic kidney disease progression.

Reference

Karter AJ, Parker MM, Finertie H, Nugent JR, Gilliam LK, Mahapatra A. Continuous glucose monitoring and long-term glycemic and kidney outcomes. JAMA Netw Open. 2026;9(10):e2637698. doi:10.1001/jamanetworkopen.2026.37698