HbA1c is the world’s most widely used test for diagnosing type 2 diabetes. New evidence shows that G6PD deficiency and other genetic variants are systematically distorting its results for Black and South Asian populations.
Every year, millions of people are diagnosed with type 2 diabetes, and for many the process begins with a single blood test. HbA1c testing is convenient, does not require fasting, and gives an integrated picture of blood sugar control over several months. For a substantial share of the global population, however, it gives a misleading one.
HbA1c (glycated haemoglobin) reflects average blood sugar levels over roughly two to three months. As red blood cells circulate, glucose attaches to haemoglobin; the longer cells have been exposed to elevated glucose, the higher the reading. A 2011 WHO consultation formalised its use as a diagnostic tool, and it is now embedded in clinical guidelines worldwide. The assumption built into this is that red blood cells behave the same way across all individuals. This assumption is wrong for a meaningful share of the global population.
The role of G6DP deficiency
Growing evidence shows that common genetic variants affecting red blood cell biology can distort HbA1c readings independently of actual blood glucose levels. One of the most striking examples involves G6PD deficiency. The condition persists at high frequency in populations from Africa, Asia, the Middle East and Mediterranean regions because carrying the variant appears to offer some protection against severe malaria, a relationship explored through host genomics research. The consequences of that evolutionary history show up, unexpectedly, in diabetes clinics.
G6PD plays a critical role in protecting red blood cells from oxidative damage. When the enzyme is deficient, red blood cells are more fragile and have a shorter lifespan than normal. Because HbA1c accumulates over the life of a red cell, people with G6PD deficiency have less time for glycation to occur, regardless of actual blood glucose. The result is an HbA1c reading that is artificially low, not because blood sugar is well controlled, but because of this unrelated feature of red cell biology.
A study published in Diabetes Care in October 2025 by Prof Barroso and colleagues brought this problem into focus. Using whole-exome sequencing linked to electronic health record data from UK Biobank (467,368 individuals) and the Genes & Health cohort (43,011), they showed that undiagnosed G6PD deficiency distorts HbA1c in ways that systematically delay diabetes diagnosis for Black and South Asian men in the UK, and very likely many more worldwide.
In the study men with G6PD deficiency were diagnosed with type 2 diabetes on average, 4.1 years older at diagnosis than men without it. That delay has measurable consequences: they had 37% higher odds of developing diabetes-related microvascular complications, including eye, kidney and nerve damage.
Lower mean HbA1c values in these men also falsely underestimate their 10-year diabetes risk, meaning that risk-stratification tools in primary care will systematically miss the very populations most likely to be affected.
The scale of the problem
The study estimated that approximately 1 in 7 Black males and 1 in 63 Asian males in the UK carry G6PD deficiency alleles, compared to fewer than 1 in 10,000 White males. Despite these numbers, fewer than 1 in 50 men with G6PD deficiency are clinically recognised. The condition is often entirely silent, and G6PD deficiency is not routinely screened for in the NHS.
The focus on men reflects the X chromosome-linked nature of the condition: because men have only one copy of the G6PD gene, they express it fully when they inherit a deficient variant, making the clinical signal cleaner and more consistent than in women. Women are largely protected if one copy is functional, though those who carry deficient variants on both chromosomes, whether the same variant or two different ones, can also be affected.
G6PD is not alone
Two preprints from March 2026 expand the picture in important and complementary directions.
A PIEZO1 missense variant carried by approximately 1 in 13 South Asians and extremely rare in other populations was associated with lower HbA1c and changes in red cell traits across over 63,000 participants in the UK, but no difference in fasting glucose or C-peptide, confirming the effect operates through red cell biology rather than glycaemia. Carriers had reduced risk of prediabetes and type 2 diabetes diagnosis, and increased risk of diabetic eye disease once diagnosed. The researchers modelled approximately 1,019 missed prediabetes diagnoses and 303 missed type 2 diabetes diagnoses per 100,000 adults over ten years. Because a prediabetes diagnosis triggers referral to the NHS Diabetes Prevention Programme, each missed diagnosis is also a missed prevention opportunity, at an estimated cost of between £970,000 and £1,390,000 per 100,000 individuals over ten years.
HbE thalassaemia trait, a common structural haemoglobin variant in South Asian populations, pulls in the opposite direction. In 43,088 British Bangladeshi and Pakistani participants, the variant was associated with increased HbA1c independently of blood glucose, higher rates of prediabetes and type 2 diabetes diagnosis, but paradoxically lower rates of diabetic eye disease and cerebrovascular disease. The authors suggest HbA1c overestimation is driving overdiagnosis, which in turn reduces apparent complication risk. It is worth noting that the direction of HbE’s effect on HbA1c depends on the laboratory assay method used; in the NHS testing context, the evidence points to HbA1c reading higher than the true level. The study also found that 3 in 4 carriers of the thalassaemia trait had no clinical diagnosis in their NHS records.
A question for diagnostics and for equity
The overall pattern is striking. G6PD deficiency and the PIEZO1 variant cause HbA1c to read too low, delaying diagnosis and increasing complication risk. HbE thalassaemia trait causes HbA1c to read too high, leading to overdiagnosis and overtreatment. Both directions of distortion fall predominantly on South Asian and Black communities. The test adopted globally for its convenience is failing the same populations in different ways.
The International Diabetes Federation’s 2024 position statement argues that many people with type 2 diabetes or prediabetes would not be detected through HbA1c or fasting plasma glucose alone, recommending a 1-hour oral glucose tolerance test as a screening tool instead. Unlike HbA1c, a glucose-based test is not affected by haemoglobin biology, red cell lifespan, or the genetic variants described above.
Where policy is heading
The research is feeding into UK health policy. The NHS Race and Health Observatory has been vocal about the fact that inequity is often embedded in clinical pathways and the evidence base itself.
The HbA1c problem illustrates this precisely: a diagnostic standard built on an incomplete understanding of human variation in red cell physiology, with consequences that have fallen unevenly across communities. The practical questions now being worked through include whether NICE guidance should recommend supplementary glucose-based testing for individuals of Black or South Asian ancestry, and whether G6PD screening should become part of routine care prior to HbA1c testing.
From the reference genome to polygenic scores, clinical tools built on data that underrepresents global human variation have consistently performed less well for the populations left out of the evidence base. Correcting that requires sustained investment in research that includes underrepresented populations from the outset, and strategic interventions that address the structural reasons why that representation has been missing.
Updating clinical guidance is necessary but the deeper fix is ensuring the next generation of diagnostic tools is not built on the same incomplete foundations. For communities that have long experienced poorer health outcomes, the fact that these distortions are now visible, quantifiable, and reaching the attention of policymakers is a meaningful step forward.
