Design: multicentre longitudinal observational study using linear mixed-effects models
Population: 1,028 adults with type 1 diabetes contributing 3,441 CGM profiles over three years
Primary outcome: longitudinal change in time in range 70-180 mg/dL, time below range and glycaemic variability by subgroup
Effect: improvement in time in range for CKD (beta = 0.228; P = 0.002) and highest TyG tertile (beta = 0.083; P = 0.001); attenuated by severe beta-cell failure (beta = -0.119; P = 0.025); older vs younger comparable (beta = 0.110; P = 0.102) with time below range 1.7% vs 3.5% (P < 0.001)

A multicentre study followed 1,028 adults with type 1 diabetes contributing 3,441 continuous glucose monitoring profiles over three years, stratified by age, chronic kidney disease, beta-cell function, BMI and triglyceride-glucose tertile. At baseline the expected pattern held: older age, CKD, severe beta-cell failure, higher BMI and the highest TyG tertile all spent more time above 180 mg/dL.

The trajectories diverged from that expectation. Older adults improved time in range comparably to younger ones and held a lower time below 70 mg/dL throughout. The CKD group improved time in range significantly without more hypoglycaemia. It was severe beta-cell failure — not age, not kidney disease — that attenuated both the gain in time in range and the reduction in time below range.

The authors' own emphasis is on the spread: CGM metrics varied widely within every subgroup. A label such as 'older' or 'CKD' predicts the baseline but not the direction of travel, which is an argument for setting targets from a patient's own CGM trajectory rather than from their risk category.

In practice that means the reflex of loosening targets for an older patient or one with CKD deserves a look at their actual data first.

Before relaxing a target for age or CKD, look at that patient's own time below range over the last three months
Residual beta-cell function is the variable most worth knowing when improvement stalls despite good engagement
Track time below 70 mg/dL as a separate outcome, not as a footnote to time in range
Review CGM trajectory across visits rather than judging from a single 14-day report
Where beta-cell failure is severe, set expectations about the ceiling on improvement early rather than after repeated intensification

Don't overread it

This was observational follow-up, not a trial — it cannot show that any intervention produced the improvement seen in the older and CKD groups.

The statistics, in plain English

The beta coefficients are from linear mixed-effects models and describe the difference in rate of change per subgroup, not an absolute difference in time in range — so beta = 0.228 for CKD means that group improved faster, not that they gained 0.23 percentage points. The older-adult comparison, beta = 0.110 with P = 0.102, is a non-significant result: it supports 'no detectable difference from younger adults', which is the point being made, but a non-significant difference is not proof of equivalence. The wide within-subgroup variation the authors report is the reason a group-level average should not be read onto an individual patient.