Swedish Study: Why Your HbA1c May Not Match Your CGM Data

A study from Sweden has uncovered an important gap in how we interpret blood sugar data for people living with type 1 diabetes — one that could directly affect how American patients and their doctors make treatment decisions.

Researchers at the University of Gothenburg's Sahlgrenska Academy analyzed data from two clinical trials — the GOLD and SILVER trials — involving 144 adults with type 1 diabetes followed over 2.5 years. The team examined whether mean glucose (MG) and time in range (TIR), two measurements commonly provided by continuous glucose monitors (CGMs), reliably predict HbA1c levels.

What the Swedish Researchers Found

The average HbA1c among participants was 8.7%, with a mean age of 44 years. While mean glucose explained about 63% of the variation in HbA1c across the group, that figure jumped to 88% when individual biological differences were factored in — meaning a significant portion of the HbA1c story is personal and not captured by population averages alone.

Critically, 1 in 10 participants had an HbA1c that deviated by more than 0.8 percentage points from what their CGM data would predict. In practical terms, this could mean a patient's CGM suggests their average glucose is in an acceptable range, while their lab HbA1c tells a different story — or vice versa.

The researchers also found that time below range (TBR) — defined as glucose below 70 mg/dL (3.9 mmol/L) — and high glucose spikes above 250 mg/dL (13.9 mmol/L) (time above range level 2, or TAR2) each influenced the relationship between TIR and HbA1c. Specifically, more low glucose episodes were linked to a lower HbA1c for the same TIR, while more severe highs were linked to a higher HbA1c.

Understanding the Numbers in US Terms

• Time Below Range (TBR): Glucose under 70 mg/dL
• Time in Range (TIR): Glucose between 70–180 mg/dL (the ADA-recommended target zone)
• Time Above Range Level 2 (TAR2): Glucose above 250 mg/dL

The American Diabetes Association (ADA) currently recommends a TIR of greater than 70% and an HbA1c below 7% for most non-pregnant adults with type 1 diabetes. This Swedish research suggests these two targets don't always move in lockstep — and that both metrics are needed for a complete picture.

Why This Matters for US Patients

Many American patients with type 1 diabetes are now using CGMs and hearing a lot about TIR as the new gold standard. This Swedish research is an important reminder that no single number tells the whole story. Here's why this research is relevant to you:

• Your biology is unique. The same CGM readings can mean different HbA1c outcomes depending on individual factors like red blood cell turnover and glucose metabolism.
• Low blood sugars can mask poor control. Frequent lows can artificially lower HbA1c, making control look better than it is — which is a safety concern.
• Talk to your care team about both metrics. Relying solely on TIR or solely on HbA1c may give an incomplete picture. Together, they offer better insight.
• Insurance and clinical decisions often hinge on HbA1c. Understanding why your lab result might not match your CGM app can help you advocate for yourself at appointments.

This Swedish research adds to a growing global conversation about moving beyond any single diabetes metric. For US patients managing type 1 diabetes, the takeaway is clear: use your CGM data and your lab HbA1c together, and ask your doctor to explain discrepancies rather than dismiss them.

Source: Sterner Isaksson, Imberg, Hirsch. "Discordance between mean glucose and time in range in relation to HbA1c in individuals with type 1 diabetes: results from the GOLD and SILVER trials." Diabetologia (2024). DOI: 10.1007/s00125-024-06151-2. PMID: 38668761.