A recent extensive study delved into the long-term effects of real-time continuous glucose monitoring (rtCGM) on adults with type 2 diabetes (T2D), particularly those at elevated risk. The investigation tracked participants over three years, revealing significant improvements in blood sugar management and a notable, albeit delayed, positive impact on certain kidney health indicators. However, the study also pointed out areas where rtCGM’s benefits require further validation through more rigorous research.
Managing glucose levels is fundamental for individuals with diabetes, as chronically elevated blood sugar can lead to severe complications, including kidney damage. Continuous Glucose Monitoring (CGM) devices, which utilize subcutaneous sensors, provide valuable data for informed diabetes management. While real-time CGM offers immediate feedback, other systems necessitate manual scanning. Previous randomized clinical trials have shown that CGM can improve glycemic control in the short term, but comprehensive data on its sustained benefits, especially concerning kidney outcomes, has been limited. This observational study aimed to bridge that gap, examining whether prolonged CGM use could preserve kidney function.
The research employed an observational comparative effectiveness design, analyzing electronic health records from Kaiser Permanente Northern California (KPNC), an integrated healthcare system. To ensure robust comparison, a target trial emulation approach was used, accounting for differences between participants who initiated rtCGM and those who did not. The study population comprised 573,663 T2D patients, focusing on adults aged 19 and above with poor glycemic control (HbA1c > 8%) or a history of hypoglycemia-related emergency visits. Participants were monitored for up to 36 months, with rtCGM initiators compared against matched controls.
The primary outcomes measured were HbA1c levels, urinary albumin-creatinine ratio (UACR), and estimated glomerular filtration rate (eGFR). UACR indicates kidney damage, while eGFR assesses kidney filtration efficiency. Statistical weighting was applied to adjust for baseline differences in patient characteristics, treatment adherence, and missing data. The study primarily focused on real-time CGM, meaning its conclusions may not directly apply to intermittently scanned systems. Sustained use was inferred from quarterly claims for rtCGM supplies rather than daily device usage.
The study encompassed 25,648 participant assignments, involving 2,771 rtCGM initiators and 22,877 controls. The analysis included 25,371 unique individuals, with a mean age of 61.5 years and 54.7% men. At baseline, the average HbA1c was 9.5%, signifying poor glycemic control. A majority (96.9%) had HbA1c above 8% in the preceding year, and 5.5% had experienced hypoglycemia-related emergency care in the last three years. Most baseline data was from 2019 or later.
rtCGM initiators typically presented with more complex clinical profiles, including longer diabetes duration, higher comorbidity burden, and poorer kidney health at baseline. Their mean diabetes duration was 17.6 years, compared to 12.6 years for controls. Insulin use was also significantly higher among rtCGM initiators (89.1%) than controls (38.8%). Statistical weighting effectively balanced these baseline disparities between the groups.
The consistent use of rtCGM correlated with swift and enduring improvements in glycemic control. One year post-baseline, the weighted mean HbA1c for rtCGM users was 7.83%, significantly lower than the 8.93% for controls—a difference of 1.10 percentage points. These improvements persisted, with a 0.90 percentage point difference at two years. Even after three years, the rtCGM group maintained a lower mean HbA1c (7.83% versus 8.58%), showing a 0.75 percentage point difference.
Initially, no statistically significant differences in UACR were noted between the groups. However, at the three-year mark, sustained rtCGM use was linked to a lower mean UACR, with values of 153 mg/g in the rtCGM group compared to 278 mg/g in controls—a reduction of approximately 125 mg/g. This indicates a later decrease in albuminuria among those who continued monitoring. Nevertheless, a lower UACR doesn't conclusively prove kidney function preservation, and the study couldn't establish a direct causal link between improved glucose control and this reduction.
Weighted eGFR estimates were already lower at baseline in the rtCGM group and remained so for the first two years. By the third year, this difference was no longer statistically significant. Overall, the study did not definitively show that prolonged rtCGM use preserved kidney filtration function during the observation period.
The study faced limitations due to substantial participant attrition and crossovers between monitoring groups. During follow-up, 44.3% of rtCGM initiators discontinued its use, while 18.7% of controls began using CGM. At three years, the UACR analysis involved only 331 rtCGM initiators and 3,890 controls, a significantly smaller sample than at baseline. Although statistical weighting adjusted for these issues, the reduced sample size introduces uncertainty in the longer-term kidney findings, particularly regarding the delayed reduction in albuminuria.
In conclusion, the sustained application of real-time continuous glucose monitoring (rtCGM) demonstrated prompt and lasting enhancements in glycemic management, coupled with a deferred decrease in urinary albumin-creatinine ratio (UACR) among high-risk adults afflicted with type 2 diabetes. However, the investigation yielded no conclusive evidence of preserved kidney function, as gauged by the estimated glomerular filtration rate (eGFR), over the three-year duration. These findings suggest that consistent rtCGM usage can refine blood glucose control and mitigate albuminuria, but it does not unequivocally confirm the technology's ability to halt the progression of kidney disease.