Same A1C Can Hide a Very Different Glucose Day

Two people can have the same A1C and very different glucose days. Time in range from a continuous glucose monitor shows the shape of those days, while A1C only shows the midpoint.

Your A1C is an average, and averages forget the shape

A1C is a lab measure of glycated hemoglobin, a protein inside red blood cells that gets coated with glucose. It reflects the previous two to three months, weighted toward the most recent few weeks. An A1C of 7% implies an average glucose around 154 mg/dL. That average is ambiguous. The same 154 mg/dL can come from a flat line near 154 or from large swings between 70 and 250. A1C cannot tell you how much of each day was spent high, low or stable.

Continuous glucose monitoring produces a reading every few minutes. That turns a three-month average into a time series. Time in range divides the day into three bands: below 70 mg/dL, 70 to 180 mg/dL, and above 180 mg/dL. It then reports the percentage of time in the middle band. That is the part A1C skips.

The same A1C can hide a very different glucose day. Time in range is the part of the day your glucose sat between 70 and 180 mg/dL.

What does time in range mean compared to A1C?

Time in range is the percentage of sensor readings between 70 and 180 mg/dL (3.9 to 10.0 mmol/L). The 70 to 180 band was set by international consensus for people with diabetes (Battelino et al., 2019), with a goal of more than 70% of the time. A1C is a slower, blood-based average of glycated hemoglobin. The two are related but not interchangeable, because one is a direct measure of recent glucose distribution and the other is a biological average.

On a lab report, A1C appears as a percentage. A CGM export from LibreView, Dexcom Clarity, Stelo, Lingo or SiBionics contains the raw glucose readings. The free HealthOS CGM report computes time in range, GMI, variability, an ambulatory glucose profile (AGP), biggest rises and overnight patterns from that uploaded export. The AGP is the daily shape: a median line with bands for the middle 50% and 90% of all readings at each hour.

One common mistake is treating the CGM's estimated A1C, often called GMI, as a lab A1C. It is a calculated estimate from average glucose. The measured lab A1C can sit higher or lower in a given person because red cell lifespan and hemoglobin variants shift the relationship.

The same A1C, two different days

On an AGP, one person might show a narrow band from 95 to 140 mg/dL for most hours, with a modest rise after dinner. Another person might ride a wide band from 60 to 230, crossing below 70 at night and above 180 after most meals. These shapes are illustrative, not lab values. A flat or narrow shape can produce high time in range, while a wide shape can produce low time in range, even if the shapes cancel to a similar average.

This sketch is illustrative. It shows why A1C alone cannot tell you what the daily glucose pattern looks like. Two profiles with different shapes can have very different time in range, highs and lows.

The numbers: 10 percentage points more TIR, 0.6 percentage points lower A1C

In an analysis of four type 1 diabetes trials with 545 adults (Beck et al., 2019), published in the Journal of Diabetes Science and Technology, the relationship between TIR and A1C was moderate, not tight. On average, a TIR of 70% corresponded to an A1C near 7%, and a TIR of 50% corresponded to an A1C near 8%. For a given TIR, there was a wide spread of possible A1C values.

An increase in time in range of 10 percentage points is equal to 2.4 additional hours per day in the 70 to 180 mg/dL band. In the Beck et al., 2019 analysis, this was associated with an average A1C fall of 0.6 percentage points. That is an average, not a promise. A given person could see a smaller or larger change.

Why the two only partly overlap

Time in range and A1C correlate because both track average glucose. They diverge because A1C is a biological average inside red cells, not a direct time series. The same average glucose can produce different TIR when the shape of the day changes. A person who spends half the day at 50 and half at 250 can have the same average as someone flat near 150, but far less time in range. A constant 160, by contrast, can produce 100% TIR with a similar average. The distribution matters.

Use A1C as the long-term anchor and TIR as the daily shape. If TIR changes while A1C is stable, the pattern is changing before the lab catches up.

Hyperglycemia metrics such as time above 180, time above 250, mean glucose and area under the curve above 180 correlate with each other at roughly 0.90 or higher. Each of them correlates with A1C only around 0.66 to 0.78. That leaves room for two people with the same A1C to have different TIR, and for a given TIR change to produce different A1C changes.

What if I don't have diabetes?

The 70 to 180 mg/dL band was set for people with diabetes. In adults without diabetes, most of the day sits inside a tighter band. One multicentre study (Shah et al., 2019) found about 96% of the time between 70 and 140 mg/dL, and a Framingham study (Spartano et al., 2024) found about 3 hours a day above 140 mg/dL. So a person without diabetes can still see post-meal rises above 140. That is normal, not a disease signal. Time in range is still useful as a shape metric, but the 70% goal is not the right pass mark for people without diabetes.

If you have LibreView, Dexcom Clarity, Stelo, Lingo or SiBionics data, drop the export into the free CGM report and you will see your time in range, GMI, variability, AGP, biggest rises and overnight patterns in the browser, no account. If you do not have a sensor, HealthOS Sensor is a month of continuous glucose for $70, no membership, and it does not ship to the US yet. Join the free waitlist for pre-launch access. HealthOS Sensor is a wellness product, not a medical device, and it does not diagnose or treat diabetes.