Our study, adding to existing international findings,1 2 10 11 is the first from the UK/Europe to document a significant, glucose-independent racial disparity in HbA1c levels for CYP from Black ethnic background. Despite equitable access to advanced diabetes management technologies and achieving similar MBG and TIR, CYP from Black heritage demonstrate higher HbA1c levels compared with both their South Asian and white counterparts. CYP from Black background also display higher proportion of hypoglycemia. SES as reflected by IMD did not influence HbA1c.

Our results, using different techniques for glucose and HbA1c measurement, corroborate and expand on existing literature from other geographic locations demonstrating a 4–5 mmol/mol higher HbA1c for CYP from Black ethnic background relative to their White counterparts, independent of MBG.12 15–17 Our study’s key strengths in comparison to other work include our use of 90-day CGM data with an iCGM system (Dexcom G6), while attaining no significant MBG difference between racial groups by using a validated structured education program.19 20 Interestingly, we did not find a difference in HbA1c at any given level of MBG in CYP of South Asian and white backgrounds. The disparity in HbA1c outcomes among white and South Asian ethnic groups noted in previous studies4 may have been due to differences in MBG associated with inequity in access to advanced diabetes technology and validated structured education. By providing equal access to CGM using a validated education program we achieved equivalent HbA1c levels through comparable MBG and TIR in the white and South Asian groups. However, our results indicate that even with equal access to CGM and education, which standardizes MBG and TIR, large disparities in HbA1c still persist between CYP from Black backgrounds compared with their South Asian and white counterparts.

Data from the Diabetes Control and Complications Trial (DCCT) were the first to highlight the clinical impact of glucose-independent effects on HbA1c, linking them to a threefold increased risk of retinopathy and sixfold increased risk of nephropathy.28 Subsequent research corroborates a strong association between glucose-independent effects on HbA1c and increased risk of microvascular and macrovascular complications for adults with T1D29 and type 2 diabetes (T2D).30 31 Furthermore, a higher HbA1c for similar glucose levels puts people with diabetes from Black ethnic backgrounds at a higher risk of chronic complications.13 32

Factors that explain racial/ethnic glucose-independent disparity in HbA1c are currently not well established or understood.33 Some suggested mechanisms include differences in red blood cell (RBC) lifespan, alterations in the glycation rate of proteins within the RBC, variation in RBC glucose transporter, or a combination of these factors.33 Nonetheless, previous studies adjusting data for RBC indices and iron status have not been able to eliminate the glucose-independent racial difference in HbA1c.34 35

The observation that CYP of Black ethnic origin exhibit higher HbA1c levels and an increased incidence of hypoglycemia compared with their South Asian and white counterparts for the same MBG levels necessitates further investigation into the clinical decision-making processes that may exacerbate these disparities. Outside of the UK, the existing literature supporting the enhanced risk of hypoglycemia in the Black population is strong, with several studies suggesting an increased risk of hypoglycemia.36–38 However, a recent CGM study by Christakis et al has demonstrated a glucose-independent racial effect on HbA1c without increased risk of hypoglycemia.15 In contrast to our study, the Black ethnic group in this study15 had a higher MBG than the white group, and the clinicians solely used CGM metrics to adjust insulin dosing and did not have point-of-care HbA1c at the time of the clinic visit.15 Due to the tendency of HbA1c to overestimate MBG levels in CYP from Black backgrounds, clinicians with access to point-of-care HbA1c testing may lean toward intensifying insulin treatment based on HbA1c readings. Such an inclination could arise even if the MBG or TIR is within the target range due to national HbA1c comparisons.5 Consequently, CYP from Black backgrounds, with comparable CGM glucose metrics, may be subject to more aggressive insulin intensification than their counterparts of other ethnicities, resulting in an elevated risk of hypoglycemia, as demonstrated in our findings. The increased incidence of hypoglycemia in the Black group is unlikely to be explained by SES as in our modeling, neither the IMD nor the need for an interpreter were significant factors. Additionally, all CYP received a validated structured education program, including one-on-one sessions for those requiring an interpreter, hence we believe that understanding of hypoglycemia management was adequate and is unlikely to be contributory.

The concerning HbA1c/hypoglycemia paradox has been demonstrated for people from Black ethnic backgrounds with T2D in the Action to Control Cardiovascular Risk in Diabetes study,39 CYP with T1D from North America38 and East Africa,40 non-Hispanic black individuals from the DCCT,41 and in a large adult data set where African American patients with T1D and T2D had higher rates of severe hypoglycemia.42 Therefore, it is important to take into account glucose data, where available, to adjust insulin for these high-risk populations rather than relying solely on HbA1c.40 Furthermore, there is a need to prioritize these high-risk groups for access to advanced diabetes technologies such as AID systems that can optimize glucose metrics without increasing the risk of hypoglycemia.43–45 Moreover, comprehending the effect of AID systems on this HbA1c/hypoglycemia paradox in CYP from Black heritage is vital, as AID algorithms use CGM data exclusively and do not consider HbA1c levels. For national benchmarking of glucose control, relying solely on HbA1c requires careful consideration.4 5 Metrics, such as TIR, may be better indicators of glycemic control and also help clinicians avoid unnecessary intensification of insulin for patients with elevated HbA1c levels with acceptable TIR. This approach can prevent the focus on achieving good standings on league tables at the expense of appropriate patient care. However, concerns have been raised about the variability in glucose results between CGM systems, which stem from differences in algorithms and comparator methods used in accuracy studies.46 This highlights the urgent need for standardization in CGM accuracy studies, including the choice of comparator methods, study procedures, and performance measurements. The IFCC working group for CGM is currently addressing this by developing study procedures.47 These study procedures are building toward an International Organization for Standardization (ISO) for CGM devices that is comparable to the ISO 15197:2003 for capillary blood glucose monitoring systems.47 This standard aims to significantly reduce variability between CGM systems and improve the effectiveness of clinical decision-making and validity of study outcomes using CGM metrics.

The study strengthens our understanding of HbA1c racial disparity by adding 90-day CGM observational data to the existing literature from a unique pediatric population in the UK. The study comprehensively assessed MBG and other glucose metrics using CGM, ensured that all participants received structured education in using glucose management tools and assessed the impact of social deprivation by IMD which was not typical of prior work. Consideration of these factors further validates data from other geographical sites in Africa and North America which have shown that HbA1c overestimates MBG for many African heritage individuals when compared with White/European people living with diabetes.

Our study has several limitations. The limited sample size and examination of only one HbA1c and 90-day CGM data pairing for each participant raise questions about the reproducibility and generalizability of the glucose-independent disparity, particularly affecting CYP with T1D of Black heritage. However, our study is among many that report this glucose-independent effect of HbA1c, generally impacting people with diabetes of Black heritage.12 15–17 Furthermore, our study could be easily replicated at other sites in the UK and Europe to further validate the generalizability of our findings. Further, our study was not designed to clarify the mechanism for this disparity nor its potential impact on development and progression of complications, which should be the focus of future studies. Additionally, the IMD does not account for family income or parental education status which might pose barriers to use of advanced technology.