What is a healthy HbA1c value?
There is no fixed 'healthy' HbA1c value: lower within the normal range appears to be better, but how low is optimal also depends on your personal circumstances. Discuss your result with your doctor if you are pregnant or have risk factors.
The official threshold for diabetes is 48 mmol/mol (approximately 6.5%). Values below that are considered 'normal', but what exactly constitutes a healthy value is not as black and white as that threshold suggests. There are also two different measurement systems, which means the same result can appear either as a percentage or as mmol/mol. This sometimes causes confusion when comparing test results.
Within the normal range, lower genuinely seems to be better. A large British study involving more than 41,000 people showed that those at the upper end of the normal range already have slightly thinner nerve cells in the retina than those with lower values. Whether this effect has long-term consequences for vision is not certain, but it does suggest that 'officially normal' is not the same as 'optimal'.
During pregnancy, this is even more pronounced. An HbA1c between 5.5 and 5.9% (still officially normal) was linked in multiple studies to more than twice the risk of preterm birth and nearly twice the risk of having an excessively large baby, compared with women who had lower values. For pre-eclampsia (a dangerous pregnancy complication involving high blood pressure), the risk was more than three times as high. This applied to women without an official diagnosis of gestational diabetes as well.
HbA1c is also not an equally reliable measure of blood sugar for everyone. The lifespan of red blood cells varies from person to person, and this influences the result. Someone with a shorter red blood cell lifespan will show a lower HbA1c, even if their blood sugar is actually higher. The reverse is also true. This means that the same number can reflect a different reality in one person than in another.
The claims are based on observational research and large database studies (including UK Biobank with 41,000+ participants and a Chinese cohort study with 5,600+ pregnant women). No intervention or randomised studies are included in the source. Causal conclusions are therefore limited.