Is Glucose Rate-of-Change a Better CGM Metric Than Time-in-Range?

This piece covers glucose rate-of-change (RoC) as a CGM metric in people without diabetes, and what one pilot study found when comparing it to standard glycemic variability metrics under a dietary intervention. It does not cover CGM use in diagnosed diabetes management.

A pilot study in 30 healthy women found that a two-week, calorie-tailored, low-glycemic-index dietary intervention measurably reduced time spent in rapid glucose rise and fall, tracked using a rate-of-change metric, while a comparable exercise intervention produced a smaller and less consistent effect on the same measure. That is a real, if preliminary, signal that rate-of-change may pick up something standard summary metrics like average glucose and time-in-range do not directly capture. It does not establish that rate-of-change is a better health metric overall, only that it responded differently to diet than to exercise in this small sample.

Two different questions a CGM can answer

Most of the numbers a non-diabetic CGM user sees are summary statistics: average glucose, time spent in a target range, how many spikes crossed a threshold. Those numbers describe where glucose sat, averaged over a day or a week. Rate-of-change is a different kind of number. It describes how fast glucose is moving at any given moment, not just where it ends up.

The distinction matters because two people can have identical time-in-range numbers while getting there very differently, one with smooth, gradual glucose movement and the other with sharp spikes and drops that happen to average out. Standard summary metrics can't tell those two patterns apart. Rate-of-change is built specifically to.

What the study actually compared

The design tested two 14-day interventions in sequence against each participant's own free-living baseline: a structured, calorie-tailored diet built around low-glycemic-index regular meals, and a separate physical activity program. Both used the same rate-of-change framework, tracking how much time glucose spent rising or falling quickly rather than just logging peaks and averages.

The dietary intervention showed a statistically real reduction in that rapid-movement time. The exercise intervention's effect on the same rate-of-change measure was smaller and did not reach the same level of statistical consistency. That is a specific, narrow finding: it says diet changed this particular metric more clearly than exercise did in this study, not that exercise has no effect on glucose stability generally, a question this design wasn't built to fully separate out.

It's also worth being direct about what a 30-person pilot study, run entirely in healthy women, can and can't establish. The authors themselves flagged real design limitations. This is a first look at whether rate-of-change is sensitive enough to detect something, not a settled answer about whether it should replace or supplement standard metrics for a general non-diabetic CGM user.

Why this connects to the broader non-diabetic CGM question

This fits into a larger pattern in recent CGM research: a real, growing body of work is testing which glucose metrics actually track meaningful physiology in people without diabetes, as covered in what a CGM actually shows without diabetes. Time-in-range and average glucose were largely built and validated for diabetes management, where the clinical stakes of sustained high or low glucose are well established. Whether the same summary metrics are the right lens for someone without diabetes, whose glucose rarely leaves a normal range in the first place, is a genuinely open methodological question, and rate-of-change is one candidate answer to it.

None of this changes the separate, already-covered question of whether wearing a CGM changes what someone eats or how they exercise in the first place, a behavioral question distinct from which number the device should be reporting, covered in does wearing a CGM change eating or exercise habits.

Common questions