Design: Multicentre longitudinal cohort, linear mixed-effects models
Population: 1,028 adults with type 1 diabetes, 3,441 CGM profiles over 3 years
Primary outcome: Change in CGM metrics (time in, above and below range) by subgroup
Effect: CKD: greater TIR improvement (β 0.228, P=0.002) without more hypoglycaemia; severe β-cell failure: smaller TIR gain (β −0.119, P=0.025)

A Korean multicentre study followed 1,028 adults with type 1 diabetes through 3,441 CGM profiles over three years, comparing subgroups thought likely to do badly: older age, chronic kidney disease, severe beta-cell failure, higher BMI and high triglyceride-glucose index.

All of these groups started with more time above 180 mg/dL. Over time, older adults improved time in range about as much as younger ones and kept low time below range. The CKD group improved time in range significantly without more hypoglycaemia. Severe beta-cell failure was the exception: it was associated with more hypoglycaemia and more variability at baseline, and with smaller gains in time in range and smaller falls in time below range.

The practical message is that age and kidney disease are not by themselves reasons to settle for looser control on CGM, whereas a patient with no residual beta-cell function may need more support to make the same progress.

Older age alone need not mean accepting a lower time in range when CGM is used well.
In CKD, time in range improved without more hypoglycaemia in this cohort.
Patients with severe beta-cell failure had more hypoglycaemia and smaller gains — check C-peptide where it would change the plan.
Within every subgroup, CGM metrics varied widely; set targets for the person, not the category.

Why it matters

It challenges the assumption that older patients and those with kidney disease will inevitably do worse on CGM.

The statistics, in plain English

The β values are slopes from mixed models: a positive β for time in range means faster improvement over time than the comparison group. A P value of 0.102 for older age means the difference from younger adults was not statistically clear — their improvement was comparable.