Time in Range as a Digital Biomarker

Time in range is the percentage of time a person's glucose remains within a target range, conventionally 70 to 180 mg/dL, or 3.9 to 10.0 mmol/L, for most adults with type 1 or type 2 diabetes. It is derived from continuous glucose monitoring, which samples interstitial glucose every few minutes, and it answers a question that glycated haemoglobin cannot: not what the average was over three months, but how much of the day was actually spent where it should be. Two people with identical HbA1c can have very different time in range, and the difference reflects variability the average conceals. The 2019 International Consensus on Time in Range set the targets now used across the field, recommending more than 70 percent of time in range for most adults with diabetes, and it is the single most frequently registered digital endpoint in the public DiMe endpoint library.

A continuous glucose monitor worn on the arm or abdomen measures interstitial glucose with a subcutaneous sensor, typically every one to five minutes, for a wear period of ten to fourteen days per sensor. Software then computes the proportion of readings falling inside the target range across the analysis window. The consensus recommends a fourteen day window with at least seventy percent of possible readings captured, because shorter windows and sparser data give unstable estimates. Sensor accuracy is described by mean absolute relative difference against a laboratory reference, and current sensors from the major manufacturers achieve single-digit values across most of the measurement range. Interstitial glucose lags blood glucose by several minutes, which matters more for detecting rapid excursions than for a percentage accumulated across two weeks.

Time in range is used across type 1 and type 2 diabetes research as a primary and secondary efficacy endpoint, in trials of insulins, non-insulin glucose lowering agents, closed-loop systems and behavioural interventions. In the public DiMe endpoint library it appears in more registered trials than any other digital measure, and in roughly half of those it is a primary endpoint, which is the strongest evidence of endpoint acceptance any measure in this library carries. Validation work has shown time in range correlates strongly with HbA1c while adding information the average omits, and it is now reported alongside HbA1c in professional standards of care. Trials commonly pair it with diabetes distress and quality of life instruments, because glucose control and the burden of achieving it move independently.

The target range is a convention, not a biological boundary, and different ranges apply in pregnancy and in older or high risk adults, so a percentage means nothing without stating the range and population. Time in range is also silent about severity and pattern: a day spent just outside the range and a day with brief extreme excursions can produce a similar figure, which is why it is always reported alongside time below and above range. Sensor accuracy varies across the glucose range and is generally poorest at the extremes. Finally, the measure requires sensor wear and data capture, and incomplete capture biases the estimate in ways that depend on when the gaps occur.

• Battelino T, et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the International Consensus on Time in Range. Diabetes Care. 2019. pubmed.ncbi.nlm.nih.gov
• Beck RW, et al. Validation of time in range as an outcome measure for diabetes clinical trials. Diabetes Care. 2019. pubmed.ncbi.nlm.nih.gov
• Danne T, et al. International consensus on use of continuous glucose monitoring. Diabetes Care. 2017. pubmed.ncbi.nlm.nih.gov
• American Diabetes Association. Diabetes technology: standards of care in diabetes 2024. Diabetes Care. 2024. pubmed.ncbi.nlm.nih.gov