Skip to content

A high HbA1c without diabetes: what else raises it

Oak
Translucent disc creatures drift in a reef lagoon, some clear and some crusted amber, beside a glowing coral sensor pad.

Yes, you can have an HbA1c in the 5.7–6.4% range with normal glucose handling. The same is true even above 6.5%. HbA1c is the fraction of hemoglobin that has been non-enzymatically glycated, which depends on both the ambient glucose concentration and how long each red blood cell has been circulating to accumulate that sugar. The value is not a glucose measurement. Anything that lengthens red cell lifespan, shifts the age distribution of your red cell population, or interferes with the assay chemistry will move the number without moving your glucose. The practical response is to measure glucose directly for two weeks and to check the red cell variables that govern the translation.

What HbA1c measures, and what it does not

Glucose reacts with the N-terminal valine of the hemoglobin beta chain to form a Schiff base, which rearranges into the stable ketoamine called HbA1c. The reaction is slow and roughly proportional to glucose exposure over the cell’s lifetime. With a mean red cell lifespan near 120 days, the reported value is a weighted average of the preceding three to four months. Roughly half the signal comes from the most recent 30 days.

Two people with identical mean glucose can differ in HbA1c by half a percentage point or more because their red cells differ in age. Longer-lived cells accumulate more glycation. That is the mechanism behind most of the confusing results discussed below.

The standard thresholds are 5.7% and 6.5% on the NGSP scale, corresponding to 39 and 48 mmol/mol on the IFCC scale. The conversion is IFCC = (NGSP − 2.15) × 10.929. Assay imprecision alone is around 2% coefficient of variation in a good lab, which is about 0.12 percentage points at an A1c of 6.0%. A single value of 5.9% and a repeat of 6.1% are the same result.

Causes of an elevated A1c that are not hyperglycemia

Iron deficiency is the most common one in people who eat well and exercise. Iron-restricted erythropoiesis slows red cell turnover, so the circulating population skews old and more glycated. Correcting iron deficiency typically lowers A1c by 0.3 to 0.5 percentage points with no change in glucose. Check ferritin, transferrin saturation, and the reticulocyte count before accepting any borderline A1c.

Anything that suppresses erythropoiesis has the same direction of effect. Vitamin B12 and folate deficiency, hypothyroidism, and splenectomy all raise A1c by aging the red cell pool. Chronic kidney disease is a special case because it pulls in both directions. Uremia and reduced erythropoietin shorten red cell survival and lower A1c, while carbamylated hemoglobin can interfere with some assays and raise the apparent value. Systematic comparison against continuous glucose monitoring in CKD shows that the agreement between HbA1c and measured mean glucose degrades in this population 1.

Hemoglobin variants are the other large category. HbS, HbC, HbE, and elevated HbF change either the glycation target or the way the analyzer separates peaks. Ion-exchange HPLC can co-elute variant peaks with the A1c peak and report a falsely high value. Immunoassays that target the glycated N-terminal peptide are less affected by some variants but not all. Boronate affinity chromatography measures total glycated hemoglobin and is the most variant-tolerant method. If your A1c is unexpectedly high, ask which method the lab used and whether a chromatogram is available.

Bariatric surgery deserves a separate mention because the postoperative state breaks HbA1c in both directions. Iron and B12 malabsorption raise it, while accelerated turnover and rapid weight loss lower it, and the resulting number tracks glucose poorly enough that alternative markers are recommended 2.

Cross-check with 14 days of CGM

The fastest resolution is to measure glucose directly. A 14-day continuous glucose monitor gives you about 4,000 readings at five-minute resolution, against which you can compute a glucose management indicator (GMI) and compare it to the lab A1c. The formula is GMI (%) = 3.31 + 0.02392 × mean glucose in mg/dL. Require at least 70% sensor capture across the 14 days before trusting the mean.

Export the raw CSV from Dexcom Clarity or LibreView and work on the timestamped glucose column directly.

import pandas as pd

df = pd.read_csv("clarity_export.csv", parse_dates=["Timestamp (YYYY-MM-DDThh:mm:ss)"])
g = df["Glucose Value (mg/dL)"].astype(float).dropna()

mean = g.mean()
gmi = 3.31 + 0.02392 * mean
cv = 100 * g.std() / mean
tir = 100 * ((g >= 70) & (g <= 180)).mean()

print(f"mean {mean:.1f} mg/dL | GMI {gmi:.2f}% | CV {cv:.1f}% | TIR70-180 {tir:.1f}%")

A gap between GMI and laboratory A1c of more than about 0.5 percentage points is the signal worth chasing. It is called the glycation gap, and it is reproducible within a person over years. A lab A1c of 6.2% with a GMI of 5.4% points at red cell biology or assay interference. The reverse pattern, low A1c with a high GMI, suggests shortened red cell survival from hemolysis or recent blood donation.

Also record coefficient of variation and time in range, because they carry information that no single average does. Two people at the same mean glucose can differ substantially in variability, and variability has been argued as an independent contributor to complication risk rather than a passive marker 3.

Second opinions from other glycation markers

Fructosamine measures total glycated serum protein and glycated albumin measures the albumin fraction specifically. Both reflect roughly the preceding two to three weeks and are independent of red cell lifespan entirely. If your A1c and GMI disagree, a glycated albumin result that agrees with the CGM confirms that hemoglobin is the problem. Both markers are distorted by altered albumin turnover. That means nephrotic syndrome, protein-losing enteropathy, and severe liver disease invalidate them.

An oral glucose tolerance test with insulin measured at 0, 30, 60, and 120 minutes gives a different view again. It separates people with normal fasting glucose but poor post-load handling from those whose A1c is high for hematologic reasons. Interpretation of a borderline OGTT belongs with a clinician who can see your full history.

Molecular markers under investigation may eventually add to this picture. Circulating microRNA panels measured years before onset have predicted incident type 2 diabetes in a high-risk cohort, independent of standard glycemic measures 4. These are research tools today rather than clinical tests.

What your genome adds

Around a third of population variance in HbA1c is attributable to genetic variation, and a meaningful part of that acts through red cells rather than glucose. The clearest example is the G6PD A− allele rs1050828, common in people of African ancestry, which shortens red cell survival and lowers HbA1c by roughly 0.8 percentage points in hemizygous males. A carrier can have a “normal” A1c while running genuinely elevated glucose.

In the other direction, variants in HFE and TMPRSS6 shift iron status, and structural HBB variants change the glycation substrate. If you hold whole-genome sequencing, query the relevant positions directly from your VCF.

bcftools view -r chr11:5227002,chrX:154536002 -i 'GT="alt"' sample.vcf.gz

Those two coordinates are the HbS site rs334 and G6PD rs1050828 on GRCh38. Extend the region list to cover HBB, G6PD, SPTA1, ANK1, and PKLR for a fuller red cell picture. A positive finding changes which biomarker you trust. It does not change treatment.

Context beyond the blood draw

Environment and behavior move A1c in ways that are easy to miss in a single reading. Long-term exposure to particulate air pollution is associated with higher insulin resistance and subclinical inflammation markers in population cohorts 5. Sleep, activity, and resting heart rate measured passively from wearables form an interconnected network with HbA1c. A change in A1c therefore often has a behavioral antecedent that shows up in data you already have 6. Body mass index remains one of the strongest single correlates of HbA1c at population scale 7.

Questions people also ask

Can you have a high HbA1c and not be diabetic?

Yes. Each of the following can raise the measured value without raising glucose.

  • Iron deficiency
  • B12 or folate deficiency
  • hypothyroidism
  • splenectomy
  • uremia
  • hemoglobin variants

The diagnostic path is to measure glucose directly with a 14-day CGM and compare the resulting GMI to the lab number.

Why is my A1C high if I eat healthy?

The most common explanations in that group are iron deficiency, an assay artifact from a hemoglobin variant, and genuine post-meal glucose excursions that a fasting glucose test never sees. A CGM distinguishes the third from the first two within two weeks. If GMI and A1c agree and both are elevated, the glucose is real and worth discussing with a clinician.

How much can stress raise A1C?

Psychological stress raises cortisol and catecholamines, which increase hepatic glucose output, so sustained stress can move glucose upward. The measured effect on HbA1c in studies of stress and diabetes distress is modest relative to diet, sleep, and body composition 8. Acute stress in the week before a blood draw has little effect on a three-month average.

What else besides diabetes raises A1c?

Chronic kidney disease, chronic alcohol use, severe hypertriglyceridemia, hyperbilirubinemia, and recent high-dose aspirin can all interfere with specific assay methods. Advanced age shifts A1c upward by roughly 0.1 percentage points per decade at constant glucose. Elevated HbA1c also appears alongside oxidative stress markers in acute cardiac events, which is one reason a single value drawn during illness should not be read as a baseline 9.

Which test should I trust if they disagree?

Trust the direct glucose measurement. CGM mean glucose and an OGTT measure the analyte of interest, while HbA1c measures a proxy that depends on red cell biology. Bring both results to a clinician rather than acting on either alone.

Footnotes

  1. Miguel Angel Cuevas-Budhart, Karen Gómez-Rosas, Rubén David Saavedra-Fernández, et al. Correlation Between Glycemic Biomarkers and Continuous Glucose Monitoring Metrics in Chronic Kidney Disease: A Systematic Review and Correlation Meta-Analysis. Journal of Clinical Medicine, 2026. https://doi.org/10.3390/jcm15166137 ↩

  2. Uchenna Esther Okpete, Haewon Byeon. Limitations of glycated hemoglobin and emerging biomarkers for diabetes care after bariatric surgery. World Journal of Diabetes, 2025. https://doi.org/10.4239/wjd.v16.i7.107928 ↩

  3. Louis Monnier, Claude Colette, David Owens. Glucose variability and diabetes complications: Risk factor or biomarker? Can we disentangle the “Gordian Knot”?. Diabetes & Metabolism, 2021. https://doi.org/10.1016/j.diabet.2021.101225 ↩

  4. Rosa Jiménez-Lucena, Oriol Alberto Rangel-Zúñiga, Juan Francisco Alcalá-Díaz, et al. Circulating miRNAs as Predictive Biomarkers of Type 2 Diabetes Mellitus Development in Coronary Heart Disease Patients from the CORDIOPREV Study. Molecular Therapy Nucleic Acids, 2018. https://doi.org/10.1016/j.omtn.2018.05.002 ↩

  5. Kathrin Wolf, Anita Popp, Alexandra Schneider, et al. Association Between Long-term Exposure to Air Pollution and Biomarkers Related to Insulin Resistance, Subclinical Inflammation, and Adipokines. Diabetes, 2016. https://doi.org/10.2337/db15-1567 ↩

  6. Elaheh Kalantari, Carlo Allocca, Rohit Ail, et al. Network analysis to identify interconnections between smartwatch-based lifestyle biomarkers and glycated haemoglobin (HbA1c). Scientific Reports, 2026. https://doi.org/10.1038/s41598-026-64281-6 ↩

  7. Rasoul Kowsar, Alireza Mansouri. Multi-level analysis reveals the association between diabetes, body mass index, and HbA1c in an Iraqi population. Scientific Reports, 2022. https://doi.org/10.1038/s41598-022-25813-y ↩

  8. Sara L. Davis, Sarah S. Jaser, Nataliya Ivankova, et al. Relationships Among Stress, Diabetes Distress, and Biomarkers in Children with Type 1 Diabetes Mellitus from Diverse Income and Racial Backgrounds. Journal of Pediatric Health Care, 2025. https://doi.org/10.1016/j.pedhc.2024.08.012 ↩

  9. Sounira Mehri, Raja Chaaba, Josef Finsterer, et al. Relevance of oxidative stress biomarkers, hemoglobin A1c, troponin-I, and angiotensin-converting enzyme metabolism to blood pressure in acute myocardial infarction: a case-control study. Redox Report, 2023. https://doi.org/10.1080/13510002.2023.2209360 ↩

Pothos builds longitudinal molecular profiles of individuals: whole-genome sequencing, single-cell RNA sequencing, and plasma proteomics, delivered as a detailed report, every data file, and stored cells. Build your profile.