If you've landed here wondering what glycated hemoglobin (HbA1c) is in diabetes, the short answer is this: it's a blood test that estimates how much glucose has attached to the hemoglobin in your red blood cells over the preceding weeks, which is why it works as an average of your glucose over that period. It remains a cornerstone of glycemic control assessment in clinical practice [13], and it's the clinical parameter most strongly linked so far to the risk of developing diabetes complications [2]. But it has an important blind spot: being an average, it doesn't reveal short-term glucose swings [2].
That's where time in range comes in — the percentage of time your glucose spends within target limits according to a continuous glucose monitor (CGM). The two indicators are closely related — in large series, the correlation between HbA1c and time in range is very good — but they're not interchangeable, and in clinical practice they often tell different stories [5, 6]. In this article you'll see how HbA1c forms, what time in range actually measures, how much discrepancy is typical between the two, what causes it (from kidneys and iron to sensor accuracy), and how to read both numbers together with your healthcare team without drawing the wrong conclusions.
What is HbA1c in diabetes
How it forms and what period it reflects
Glucose circulating in the blood binds slowly and steadily to hemoglobin in red blood cells. The more glucose circulating, and the longer it persists, the higher the proportion of glycated hemoglobin, expressed as a percentage (%) and also in mmol/mol: for example, 7.5% equals 58 mmol/mol [3].
Because red blood cells live for several weeks, the result doesn't depend on what you had for breakfast yesterday, but on the preceding weeks as a whole. In fact, one of the biggest reasons HbA1c and sensor metrics don't line up is that recent glucose can differ from glucose levels 8 to 10 weeks before the blood test [7]. And anything that alters red blood cell turnover changes the result without your actual glucose having changed [8, 13].
Why it remains the clinical reference
HbA1c is, as of today, the best parameter available in the clinic to estimate complication risk [2]. It's also used as a criterion in research and trials: for example, in a 2024 study in insulin-treated type 2 diabetes, participants were required to have an HbA1c ≥ 7.5% (58 mmol/mol) to enroll, with a median baseline of 8.3% [3].
That solid track record coexists with an increasingly explicit acknowledgment of its limits. Conventional standards of care — capillary blood glucose monitoring and HbA1c — have enabled glycemic control to be achieved, but they have recognized limitations [14]. And recent reviews warn that discordance between HbA1c and actually measured glucose values is common in everyday practice [13].
HbA1c, GMI, and average glucose aren't the same number
Here it's worth separating three concepts that often get mixed up:
- Laboratory HbA1c: measured in blood, reflecting hemoglobin glycation [13].
- Sensor average glucose: the arithmetic mean of all CGM readings over a period.
- GMI (Glucose Management Indicator): an estimate of HbA1c calculated from sensor average glucose, developed in a specific cohort of people with diabetes [9].
GMI resembles an HbA1c, but it isn't one. Discrepancies between laboratory-measured HbA1c and GMI are a relevant and well-documented clinical issue [8].
What time in range is and what it adds to HbA1c
The target range and the metrics that accompany it
With a CGM sensor you can calculate the percentage of time spent within, above, and below your glucose target range, which supports more person-centered care [14]. The range commonly used for time in range in adults is 70–180 mg/dL (3.9–10.0 mmol/L) [11].
Time in range never stands alone. Alongside it you'll see time above range, time below range, and glycemic variability — metrics that improve together when CGM is used instead of capillary glucose monitoring alone [17]. That dynamic snapshot is exactly what an average like HbA1c can't offer [2].
Why an international consensus was needed
For years, the barrier to routine CGM use wasn't so much sensor accuracy as the lack of clear, agreed-upon glycemic targets that both healthcare teams and people with diabetes could work toward [1]. That's why an international panel of experts (the ATTD consensus) defined unified clinical targets for interpreting and communicating CGM data across different populations with diabetes [1].
That work turned time in range into a common language: a number that can be understood in a ten-minute appointment and that you can check yourself in your sensor's app any day of the week [1, 14].
How HbA1c and time in range relate to each other
The relationship between the two indicators is close and inverse: the more time in range, the lower the HbA1c. In the classic analysis by Vigersky and McMahon, the correlation between HbA1c and percentage of time in range was excellent (R = −0.84; R² = 0.71) [5].
The practical rule derived from that work is easy to remember: for every absolute 10% change in time in range, HbA1c changes by about 0.8% (9 mmol/mol) [5]. In other words, going from 50% to 60% time in range is usually accompanied by a noticeable improvement in lab results. The authors themselves suggested that time in range could eventually become the preferred metric for assessing individual glycemic control and predicting complication risk [5].
Now for the important nuance: an R² of 0.71 means time in range explains roughly seven out of every ten units of variation in HbA1c between people [5]. The rest — almost a third — depends on other things: individual biology, glycemic variability, sensor characteristics, and lab characteristics. That unexplained portion is exactly where the two numbers stop telling the same story.
| Laboratory HbA1c | Time in range (CGM) | |
|---|---|---|
| What it measures | Glycation of hemoglobin in red blood cells [13] | % of time within, above, and below target [14] |
| Period it reflects | Previous weeks, tied to red blood cell turnover [7, 8] | The days you've worn the sensor [11] |
| Captures hypoglycemia | No: it's an average and doesn't capture short-term fluctuations [2] | Yes, through time below range [17] |
| Risk evidence | The best clinical parameter available for complication risk [2] | Associated with lower risk of albuminuria, retinopathy, and mortality [2, 16] |
| Main interferences | Kidneys, iron, hemoglobin variants, red blood cell turnover [13] | Sensor accuracy and calibration, differences between devices [8, 11] |
Why HbA1c and time in range sometimes tell different stories
What an average can't see
An HbA1c of 7% can correspond to calm days with few swings, or to a rollercoaster of hypoglycemia and hyperglycemia that cancel out in the average. By definition, HbA1c doesn't reveal short-term glucose fluctuations [2]. Time in range, together with time below range and variability, does make that difference visible [17].
This is the classic "good HbA1c, poor control" situation: seemingly fine lab numbers that hide repeated overnight hypoglycemia. Recognizing this matters, because failing to detect discordance between HbA1c and measured glucose can lead to treatment being intensified when it shouldn't be, increasing hypoglycemia risk and causing distress for the person with diabetes [13].
What shifts HbA1c without your glucose actually changing
HbA1c depends on the life and turnover of red blood cells, so several situations shift it up or down without your actual glucose being different. Common causes of discordance described include [13]:
- Iron deficiency.
- Chronic kidney disease.
- Alterations in red blood cell turnover.
- Hemoglobin variants.
- Rapid glucose changes in the preceding weeks.
This isn't theoretical. In a real-world analysis of 641 visits, discordance between HbA1c and GMI was greater in people with advanced chronic kidney disease (estimated glomerular filtration rate < 60) [6]. And in another cohort, discordance increased with higher body mass index and with a diagnosis of type 2 diabetes [7].
There's also biological variability in glycation: some people's hemoglobin glycates more ("high glycators") and others' less, for the same average glucose [10].
What depends on the sensor and the device
The other side of the equation matters too. Discordance can originate from CGM accuracy itself, sensor calibration, and physiological factors like red blood cell turnover [8].
And not all systems give the same number. When comparing two CGM systems worn simultaneously for 14 days in adults with type 1 diabetes, time in range 70–180 mg/dL (3.9–10.0 mmol/L) differed significantly between devices (62.8% versus 58.8%) [11]. The greatest discordance between systems was seen precisely in the hypoglycemic range [11], which is where decisions matter most.
How much discordance is typical
Numbers help calibrate expectations. In the analysis of 641 visits, only 11% of patients had an HbA1c–GMI discordance of less than 0.1%, while 50% had differences ≥ 0.5% and 22% had differences ≥ 1% [6]. In another cohort of 278 people using intermittently scanned monitoring, the mean absolute deviation between GMI and HbA1c ranged between 0.6% and 0.65% [7].
The authors' conclusion is cautious and applies to you too: HbA1c may not reflect average glucose as faithfully as once assumed [6]. To help organize the issue, a quick checklist has been proposed covering the main sources of discordance between HbA1c and GMI [8].
What changes depending on diabetes type and age
Type 1 diabetes
In type 1 diabetes, the evidence favoring CGM is strong: it's associated with reduced HbA1c and hypoglycemia, along with increased time in range [4]. Here, time in range mainly contributes information lab tests can't give, such as time below range and the times of day when problems cluster [17].
At the same time, in type 1 diabetes, longitudinal data are lacking: although time in range has proven the most consistent metric, more studies are needed to confirm its relationship with complications, especially in this population [2].
Type 2 diabetes
In type 2 diabetes the evidence is somewhat less robust, but points in the same direction [4]. A meta-analysis in adults with type 2 diabetes found that, compared with capillary glucose monitoring, CGM was associated with a mean HbA1c difference of −3.43 mmol/mol (−0.31%), along with a time in range increase of +6.36%, less time below range (−0.66%), less time above range (−5.86%), and lower glycemic variability (−1.47%) [17].
In a 12-month randomized trial in people with insulin-treated type 2 diabetes and insufficient control, CGM use significantly improved both time in range (between-group difference of 15.2%) and HbA1c (−0.9%, equivalent to −9.4 mmol/mol) compared with capillary glucose monitoring [3]. Recent guidelines issue a strong recommendation in favor of CGM in insulin-treated type 2 diabetes, including basal insulin alone, and also in people not treated with insulin, particularly those with HbA1c ≥ 7% [15].
One detail worth noting: discordance between GMI and HbA1c was greater in people with type 2 diabetes and higher body mass index [7], so in this group it's best not to read either number in isolation.
Children, adolescents, and people without diabetes
In pediatric populations with type 1 diabetes, HbA1c–GMI discordance calculated over 12 weeks was less than 0.1 in 24.8% of cases, ≥ 0.5 in 33.9%, and ≥ 1.0 in 9.2% [12]. In other words, in more than a third of children and adolescents the difference can be clinically relevant, something that should be kept firmly in mind when comparing both indices in daily practice [12].
In people without diabetes who use a sensor out of curiosity or for health tracking, the mismatch is even greater: mean GMI was 0.59% higher than laboratory HbA1c, and a discordance of 0.4% or more appeared in 71% of participants, compared with 39% of the original cohort in which GMI was developed [9]. If you don't have diabetes and your sensor shows a "high" GMI, that doesn't equal a diagnosis: diagnosis is established by your doctor using validated tests.
Which indicator better predicts complication risk
This is the underlying question, and the honest answer is that both contribute. HbA1c remains the best clinical parameter available for assessing the risk of developing complications [2]. But a systematic review of CGM-derived metrics found that greater time in range is associated with lower risk of albuminuria, retinopathy, cardiovascular mortality, all-cause mortality, and abnormal carotid intima-media thickness, and that time in range was the most consistent measure [2].
In type 2 diabetes, another systematic review concluded that a 10% increase in time in range was associated with reduced albuminuria, reduced severity of diabetic retinopathy, and lower prevalence of diabetic peripheral neuropathy and cardiac autonomic neuropathy [16]. Even so, the authors themselves highlight the scarcity of data directly assessing the association between time in range and microvascular complications in this population [16].
There are also results that call for caution with sophisticated interpretations. In a study of adults with type 1 diabetes, those with a GMI/HbA1c ratio < 0.9 (so-called "high glycators") had more retinopathy in raw terms (31.3% versus 23.1%), but the ratio wasn't an independent predictor after adjusting for other factors; absolute HbA1c remained the strongest glycemic predictor of retinopathy [10]. In other words: the ratio between the two indicators doesn't yet work as a standalone prognostic tool.
A key piece of the puzzle is also missing: there are no randomized trials directly comparing treatment strategies guided by time in range versus those guided by HbA1c, nor robust evidence on what level of discordance should change a clinical decision. This is territory where caution and conversation with your care team matter more than any fixed rule [2, 16].
How to read your two numbers in real life
An example you'll see often
Imagine two people with the same HbA1c of 7.4%. The first has 70% time in range and barely any time below range. The second has 55% time in range, 6% time in overnight hypoglycemia, and frequent post-meal spikes. The lab result doesn't distinguish between them, because an average doesn't capture short-term fluctuations [2]; the sensor report does, because it breaks down time within, above, and below target [14].
The second person doesn't necessarily need "more treatment": they need to adjust how things are distributed throughout the day. Intensifying treatment without looking at hypoglycemia is exactly the mistake that reviews on discordance warn against [13].
What to check when the two numbers don't add up
Use this list as a script for your next appointment, not as a basis for changing anything on your own [8, 13]:
- Sensor coverage: how many days did you wear the sensor during the analyzed period, and how many hours per day did it record?
- Time window: does the sensor data correspond to the same weeks as the lab test, or did you improve (or worsen) right before the test? [7]
- Device: have you switched CGM systems? Different sensors can give different times in range for the same person [11].
- Red blood cell health: anemia, iron deficiency, hemoglobin variants, recent transfusions, or pregnancy [13].
- Kidney function: discordance is greater with advanced chronic kidney disease [6].
- Hypoglycemia: is there time below range that didn't show up in capillary glucose checks? [17]
How to track this in your own data
For the numbers to mean something, they need context. Review your sensor report by time of day rather than just looking at the overall percentage: most patterns (nighttime, morning, after-meal periods) repeat and are better corrected separately [1, 14].
Keeping a log helps turn curves into decisions. If you record meals, exercise, medication, and symptoms in an app like GlucoBeat, each event gets linked to its later effect on glucose, making it easier to see why a particular day went out of range. Bring two or three concrete examples to your appointment; they're usually more useful than a whole month of unannotated data.
When to talk to your healthcare team
Ask for a review if your HbA1c and time in range have told opposite stories for two periods in a row, if your sensor's GMI consistently differs from your lab result by 0.5% or more [6], if time below range shows up that you weren't noticing, or if you have anemia, kidney disease, or any condition that alters red blood cells [13].
Don't adjust doses or regimens on your own based on a discrepancy between two numbers: any treatment change should be agreed with your care team, who can assess whether the problem lies in the lab test, the sensor, or the glycemic pattern [8, 13].
Frequently asked questions
What is HbA1c and what is it used for?
It's the fraction of hemoglobin to which glucose has bound, expressed in % and mmol/mol (for example, 7.5% equals 58 mmol/mol) [3]. It serves as a reference for glycemic control over the preceding weeks and is the best clinical parameter available for estimating complication risk, although it doesn't reflect short-term fluctuations [2, 13].
Is time in range better than HbA1c?
They don't compete: they complement each other. HbA1c retains the largest body of prognostic evidence [2], while greater time in range is associated with lower risk of albuminuria, retinopathy, and mortality, and is the most consistent CGM metric [2]. It has been suggested that time in range could become the preferred metric, but longitudinal studies are still lacking, especially in type 1 diabetes [2, 5].
Why doesn't my sensor's GMI match my lab result?
Because they measure different things and each has its own interferences: sensor accuracy and calibration, red blood cell turnover, and other physiological conditions [8]. It's very common: in a real-world analysis, 50% of people had differences ≥ 0.5% and 22% had differences ≥ 1% between HbA1c and GMI [6].
Can HbA1c give a misleading value?
Yes. Iron deficiency, chronic kidney disease, alterations in red blood cell turnover, hemoglobin variants, and rapid glucose changes can all cause HbA1c to not correspond with measured glucose [13]. Failing to recognize this can lead to inappropriately intensified treatment, increased hypoglycemia risk, and distress [13].
How much should my HbA1c rise if I lose time in range?
Is time in range useful if I have type 2 diabetes without insulin?
Recent guidelines strongly recommend CGM use also in people with type 2 diabetes not treated with insulin, particularly those with HbA1c ≥ 7% [15]. Overall, in adults with type 2 diabetes, CGM was associated with more time in range (+6.36%) and lower HbA1c (−0.31%) compared with capillary glucose monitoring [17].
Key takeaways
- HbA1c estimates your average glucose over the preceding weeks; time in range describes how that glucose is distributed throughout the day.
- Both indicators correlate well, and about 10 points of time in range equal, on average, about 0.8 points of HbA1c, but the relationship isn't exact person to person.
- Discordance between HbA1c and sensor GMI is common: half of people show differences of half a point or more.
- Kidneys, iron, hemoglobin variants, and red blood cell turnover alter HbA1c; accuracy, calibration, and sensor model alter time in range.
- A seemingly good HbA1c can hide hypoglycemia: always check time below range.
- In type 2 diabetes with overweight, in children and adolescents, and in people without diabetes, discrepancy between both indicators tends to be greater.
- Bring both numbers, along with your notes on meals and activity, to your appointment: treatment decisions are made with your healthcare team, never based on a single isolated figure.
References
- Battelino T et al. (2019). Clinical Targets for Continuous Glucose Monitoring Data Interpretation: Recommendations From the International Consensus on Time in Range. Diabetes Care.
- Yapanis M et al. (2022). Complications of Diabetes and Metrics of Glycemic Management Derived From Continuous Glucose Monitoring. The Journal of Clinical Endocrinology and Metabolism.
- Lind N et al. (2024). Comparing Continuous Glucose Monitoring and Blood Glucose Monitoring in Adults With Inadequately Controlled, Insulin-Treated Type 2 Diabetes (Steno2tech Study): A 12-Month, Single-Center, Randomized Controlled Trial. Diabetes Care.
- Lin R et al. (2021). Continuous glucose monitoring: A review of the evidence in type 1 and 2 diabetes mellitus. Diabetic Medicine.
- Vigersky RA, McMahon C (2019). The Relationship of Hemoglobin A1C to Time-in-Range in Patients with Diabetes. Diabetes Technology & Therapeutics.
- Perlman JE et al. (2021). HbA1c and Glucose Management Indicator Discordance: A Real-World Analysis. Diabetes Technology & Therapeutics.
- Fellinger P et al. (2022). HbA1c and Glucose Management Indicator Discordance Associated with Obesity and Type 2 Diabetes in Intermittent Scanning Glucose Monitoring System. Biosensors.
- Lenters-Westra E et al. (2025). Managing discordance between HbA1c and glucose management indicator. Diabetic Medicine.
- Shah VN et al. (2023). Discordance Between Glucose Management Indicator and Glycated Hemoglobin in People Without Diabetes. Diabetes Technology & Therapeutics.
- Sager-La Ganga C et al. (2025). The GMI/HbA1c ratio does not independently predict diabetic retinopathy in adults with Type 1 Diabetes. Scientific Reports.
- Zhou Y et al. (2022). Discrepancies in glycemic metrics derived from different continuous glucose monitoring systems in adult patients with type 1 diabetes mellitus. Journal of Diabetes.
- Piona C et al. (2022). Evaluation of HbA1c and glucose management indicator discordance in a population of children and adolescents with type 1 diabetes. Pediatric Diabetes.
- Burroughs B (2026). When HbA1c Lies: Recognizing and Managing HbA1c-Glucose Discordance in Clinical Practice. Southern Medical Journal.
- Mohan V et al. (2023). Expert Consensus Recommendations on Time in Range for Monitoring Glucose Levels in People with Diabetes: An Indian Perspective. Diabetes Therapy.
- Giaccari A et al. (2026). Evidence-based recommendations for the use of continuous glucose monitoring in type 2 diabetes: the Italian guidelines. Acta Diabetologica.
- Raj R et al. (2022). Time in range, as measured by continuous glucose monitor, as a predictor of microvascular complications in type 2 diabetes: a systematic review. BMJ Open Diabetes Research & Care.
- Jancev M et al. (2024). Continuous glucose monitoring in adults with type 2 diabetes: a systematic review and meta-analysis. Diabetologia.
Reviewed by Dr. Julie Verzura before publication. Every fact in the article links to the clinical guideline, study or official body that supports it. Reviewed on 21 Nov 2026.
Informational content. It does not replace advice from your doctor or healthcare team.