Common HbA1c Assay May Miss Diabetes in People with Sickle Cell Trait

HbA1c is widely used for diagnosis and monitoring, but certain methods may produce biased results in this population. These discrepancies can underclassify prediabetes or miss diabetes. A new study shows that an immunoturbidimetric HbA1c assay can substantially underestimate values compared with high-performance liquid chromatography in this group.

Researchers at Korle Bu Teaching Hospital in Accra, Ghana, evaluated HbA1c measurement by immunoturbidimetry against high-performance liquid chromatography (HPLC) in adults with and without sickle cell trait. The work, presented on July 28, 2026, at the ADLM 2026 meeting in Anaheim, California, addresses whether method-specific differences could alter diabetes classification thresholds for individuals carrying hemoglobin variants. Investigators emphasized the need for monitoring strategies that account for variant hemoglobins to avoid missed diagnoses.

Immunoturbidimetry, a common laboratory assay used to quantify specific proteins in a sample, was compared with HPLC, a technique that separates and measures individual components in liquids. Initial HbA1c measurements were performed on an HPLC platform capable of automatically detecting hemoglobin variants. Samples suspected of containing variants were then reassessed with immunoturbidimetric assays.

The retrospective analysis included 1,283 consecutive HbA1c tests obtained from patients aged 18 years and older in January and February 2026 at a tertiary hospital in Accra. Hemoglobin electrophoresis confirmed sickle cell trait. Of the cohort, 1,027 individuals were categorized as HbAA (normal), 198 as HbAS (sickle cell trait), and about 58 with other variants were excluded from further analyses. Patient classification was compared using standard HbA1c thresholds for prediabetes (5.7%) and diabetes (6.5%).

Results showed minimal bias between methods in the HbAA group. In contrast, among those with HbAS, immunoturbidimetry significantly underestimated HbA1c relative to HPLC (5.1% vs. 5.9%). Using the immunoassay, 11.1% of HbAS participants were labeled prediabetic versus 34.5% by HPLC, and none met the diabetes threshold by immunoassay compared with 4% by HPLC.

The authors indicated that reliance on immunoturbidimetric assays alone could lead to underclassification of prediabetes and failure to detect diabetes in people with sickle cell trait. They suggested considering HPLC or reflex testing algorithms that account for hemoglobin variants, and noted that broader awareness of method limitations could inform laboratory protocols and policies in West Africa.

“We have a high prevalence of sickle cell disease — about 2% of all newborns — and sickle cell trait in about 30% of the Ghanaian population. However, the influence of hemoglobin variants on HbA1c testing hasn’t been well characterized in our region. Our study indicates that relying on immunoassays alone likely contributes to a significant underdiagnosis of diabetes. We recommend adopting other laboratory methods that factor in hemoglobin variants, like HPLC or mandatory reflex testing algorithms, for accurate diabetes assessments in high-risk populations,” said Dr. Elikem Kumahor, a specialist laboratory physician at Korle Bu Teaching Hospital in Accra, Ghana.

Related Links
Korle Bu Teaching Hospital