Haemoglobin types and variant interference with HbA1c and its association with uncontrolled HbA1c in type 2 diabetes mellitus

Abstract

Diabetes mellitus is among the leading global health concerns, causing over 1.5 million deaths alongside other significant comorbidities and complications. Conventional diagnosis involves estimating fasting, random blood glucose levels and glucose tolerance test. For monitoring purposes, long-term glycaemic control has been achieved through the measurement of glycated haemoglobin (HbA1c) which is considered reliable and preferred tool. However, its estimation could be affected by haemoglobin types like HbA0, HbA2, and HbF concentrations whose magnitude remains unclear as well as other haematological parameters. As such, the current study determined the association between HbA1c and haemoglobin types and determined correlation between haemoglobin types and haematological parameters among patients with type 2 diabetes mellitus (T2DM) compared to healthy non-diabetic participants. In this cross-sectional study, participants [n = 144 (72 per group), ages 23–80 years] were recruited and the desired parameter measured. HbA1c and other Haemoglobin variants were measured using ion-exchange high-performance liquid chromatography (HPLC) by the Bio-Rad D-10 machine (Bio-Rad Laboratories, Inc). Haematological parameters were measured using the Celtac G MEK-i machine (Nihon Kohden Europe). SPSS version 27 (IBM Corporation, Chicago, Illinois, United States) was used for the analysis. Chi-square (χ²) analysis, Mann-Whitney U test, Binary logistic regression and Pearson correlation were used to determine the differences between proportions, compare laboratory characteristics, associations and correlations respectively. With non-diabetics as the reference group, HbA1c was associated with increased HbA0 [OR = 1.509, 95% CI = 1.020–1.099, p = 0.003] and increased HbA2 [OR = 3.893, 95% CI = 2.161–7.014, p = 0.001]. However, there was no significant association between HbA1c and HbF [OR = 2.062, 95% CI = 0.873–4.875, p = 0.099]. Further, haematocrit (HCT) had a negative correlation with HbAO and a positive correlation with HbAS in participants with controlled diabetes. Mean cell volume (MCV) and mean cell haemoglobin (MCH) had a negative correlation with HbF. MCHC (mean cell haemoglobin concentration) had a negative correlation with HbA2 in participant with uncontrolled diabetes. The study concluded that levels of various haemoglobin types should be considered while monitoring glycaemic control through HbA1c. Additionally, MCHC should be considered in individuals with high concentration of HbA2 among T2DM patients while interpretating results for HbA1c.

Background

Type 2 Diabetes mellitus (T2DM), alongside its associated complications, presents a growing health concern worldwide. Generally, T2DM is composed of various heterogenous disorders with eventual increase in blood glucose concentration [1]. As a metabolic disorder, DM occurs as a result of impaired insulin production by the pancreatic beta cells or peripheral resistance to insulin [2]. The pathophysiology of diabetes mellitus entails chronic disruptions in carbohydrate, fat and protein metabolism primarily that lead to hyperglycaemia, affecting the normal functioning of various body organs [3]. The disease has been associated with various complications, among them cardiovascular disorders such as coronary heart disease and stroke, non-vascular diseases such as cancer, mental and nervous system complications, infections, as well as liver disease [4]. The prevalence of T2DM has been growing steadily, having a close association with tremendous economic development, urbanisation and widespread adoption of modern sedentary lifestyles [5].

The standardised national prevalence of DM in Kenya is at 3.6% [6]. The estimated prevalence of DM across Western Kenya region is relatively higher at 5.1% [7], indicating a growing burden of DM across the region. Bungoma, which is part of Western Kenya, has been shown to register approximately 2349 new cases of diabetes mellitus in a year making it an ideal study site [8].

This study aimed at investigating the relationship between HbA1c levels and other haemoglobin types/variants that include HbAA, HbAS, HbSS, and HbFF. By understanding the relationship between the aforementioned parameters, a more focused interpretation would be important to improve the diagnostic approaches for type 2 diabetes mellitus. There has been no existing local data highlighting HbA1c interferences as a possible pitfall in monitoring of type 2 diabetes mellitus, a gap that this study aimed to fill. However, findings from other studies in published literature as described in this section and literature hypothesised possible interferences based on local population characteristics, forming the background for the current study.

Previously, studies have focused on the relationship between HbA1c, haemoglobin types/variants and haematological parameters to understand their implications on health. In fact, one of the key instruments in detecting the existence of haemoglobin variants is by measuring the red blood cell indices through a routine complete blood count [9]. The occurrence of haemoglobin variants is common within the tropical regions due to the chronic presence of other diseases such as malaria [10]. Although the relationship between Hb variants and haematological parameters has been enumerated by other studies, the current study focusses on the association of the aforementioned parameters under the influence of a metabolic disorder- diabetes mellitus. Thus, the study sought to address the gap in the current literature on the influence of type 2 DM as a metabolic disorder on Hb types/variants.

Despite numerous studies determining the role of HbA1c in assessing diabetes as well as its implications on other conditions in both diabetic and non-diabetic patients, there is limited data on the association between HbA1c and haemoglobin variants, especially in Africa. A related study investigated the impact of Hb variants such as HbD, Hb Louisville, Hb Las Palmas, Hb N-Baltimore, Hb Porto Alegre and HbS on HbA1c [11]. However, the authors found inconsistent correlations especially based on the methods used and the specific Hb variants and recommended further investigations within the context of locally occurring haemoglobin, which forms partly the aim of the current study.

Methods

Study site

The study was conducted at the Bungoma County Referral Hospital. The hospital is the main public health facility within Bungoma County, serving the populations within the county and neighbouring counties. The county is situated in Western Kenya, lying between latitudes 00 28’ and 10 30’ North of the Equator and longitudes 340 20’ and 350 15’ on the East of Greenwich Meridian [12]. Bungoma county’s economy is largely dependent on agriculture, while the most significant health concerns include malaria, tuberculosis, pneumonia, diarrhoeal diseases, hypertension, diabetes mellitus, and accidents [13]. The county has a total population of 1.7 million people, out of which over 3,000 are estimated to have diabetes mellitus [14]. Bungoma County Referral Hospital was the preferred study site owing to its capacity as a referral centre to handle diabetic patients from across the county, which gives a fair representation for all the nine sub-counties in the county into the study.

Study design

This was a comparative cross-sectional study involving both diabetic and non-diabetic patients aimed at establishing the relationship between HbA1c and various haematological parameters as well as haemoglobin types/variants.

Study population

The study targeted patients (n = 72) with type 2 diabetes mellitus attending clinics and/or sent to the medical laboratory for prognostic tests. An additional non-diabetic group (n = 72) was also recruited for comparison purposes (control).

Inclusion criteria

The study included all patients with type 2 diabetes mellitus attending the hospital clinics and/or sent to the laboratory for diagnostic or prognostic tests. In addition, a similar proportion of non-diabetic participants was included for comparison purposes.

Exclusion criteria

The study excluded patients with underlying endocrinological diseases that are known to affect glucose metabolism; particularly hypothyroidism and hyperthyroidism, haematological diseases, systemic diseases, pregnancy, nervous system degenerative diseases, malignancy and anaemia of chronic diseases. Monitoring for the underlying diseases was done through the electronic patient cards available through the hospital electronic health management system (EHMS).

Ethical considerations

Ethical approval was obtained from Maseno University Scientific Ethical Review Committee (MUSERC, approval number MUSERC/01231/23). Permission to conduct research granted by the National Commission for Science, Technology and Innovation (NACOSTI, approval number NACOSTI/P/23/28581). The purpose of the study was clearly explained to potential participants in a language they can understand and enrolment into the study was based on free will to participate. Participants were also at liberty to exclude themselves from the study whenever they deemed fit to do so without dire consequences. An informed, written and voluntary consent was obtained from participants before recruitment into the study. Strict confidentiality was maintained and all personal identifiers were removed from data during analysis to enhance anonymity. Both electronic and physical data was safely stored and none was relayed to any third party without the participants’ express consent to maintain high level of confidentiality. Privacy of study-related data was ensured by handling and storage of all patient information in a password-protected electronic laboratory information management system (LIMS) with strict access-control.

The benefits of participation in the study included identification, immediate treatment and referral for further treatment where necessary for the treatment or management of diabetes and haemoglobin disorders. No extra costs were charged to the participants because of this study. The study limitation was that it mainly focused on patients attending diabetic clinics, which may increase chances of missing potential participants that present in other hospital departments. Potential bias included missing out on potential participants that are key to the study, which was minimised by ensuring a highly representative sample is obtained through a well-coordinated random sampling.

Data collection

Recruitment and sample collection

Probability sampling was utilised in selecting the sample, in which participants were selected randomly and independently. A simple random sampling technique was employed for both the participants and control groups, where individuals were randomly selected until the sample size of 72 for each group was attained. For patients who consented to participate in the study, 4mls of venous whole blood was collected in sodium EDTA vaccutainers.

Laboratory procedures

HbA1c measurement

HbA1c levels were estimated using the Bio-Rad D-10 Haemoglobin Machine (Bio-Rad Laboratories, Inc). 4mls of venous whole blood was collected in sodium EDTA vaccutainers. Sample information was then entered onto the machine software, an appropriate program (Hb test panel) selected, and samples loaded into the machine. Using ion-exchange high-performance liquid chromatography (HPLC), the HbA1c levels were enumerated.

Haemoglobin type determination

Haemoglobin types were determined using the Bio-Rad D-10 Machine (Bio-Rad Laboratories, Inc). 4mls of venous whole blood were collected into sodium EDTA vaccutainers. Using ion-exchange high-performance liquid chromatography (HPLC), the levels of various haemoglobin types were enumerated and obtained as printout reports. The reports were interpreted based on the reference ranges established by Bungoma County Referral Hospital for the local population and listed in the Standard Operating Procedure.

Complete blood count

Haematological parameters were measured using Celtac G MEK-9100 K (Nihon Kohden Europe). The machine uses the DynaHelix Flow technology in which WBCs RBCs, and platelets are aligned for high impedance counting precision by utilising the highly advanced hydrodynamic-focused sheath before the cells pass through the aperture. 3-4mls of blood samples from participants were collected in EDTA anticoagulated vacutainers. A full haemogram report (Complete Blood Count) was obtained and the intended haematological parameters (RBC count, Hb, HCT, MCV, MCH, MCHC) extracted from the report. Interpretation of the report was based on the normal reference values as established by Bungoma County Referral Hospital Laboratory and listed in the Standard Operating Procedure.

Data analysis

The Statistical Package for Social Science (SPSS V.27) (IBM Corporation, Chicago, Illinois, United States) was used for the analysis. Chi-square (χ²) analysis was used to determine differences between proportions. Mann-Whitney U test was used to compare laboratory characteristics between diabetics and non-diabetics. Association between HbA1c and age was determined using binary logistic regression analysis while controlling for haemoglobin variants. Using non-diabetics as the reference group, association between HbA1c and haemoglobin types in Type 2 Diabetes Mellitus was determined using multivariate logistic regression analysis controlling for haemoglobin variants. Association between red cell parameters and haemoglobin types/variant in participants with controlled and uncontrolled HbA1c was determined using Pearson correlation. p < 0.05 was considered statistically significant.

Results

General, Demographic and Laboratory characteristics of the study participants

Diabetic patients (n = 72) and a non-diabetic group (n = 72) were recruited into the study as participants and controls respectively. For the diabetics, 31.9% were males and 68.1% females while for the non-diabetics, 27.8% were males and 72.2% were females. The proportions of haemoglobin types (HbA0, HbA2 and HbF) with regard to the diabetic status (i.e., diabetics and non-diabetics) was significantly comparable between diabetic and non-diabetic participants. Diabetic participants had lower HbA0 [median (IQR); 79.9 (5.9)] relative to non-diabetics [median (IQR); 86.4 (3.2), p = 0.001]. Furthermore, diabetic patients had lower HbA2 [median (IQR); 2.3 (1.2)] than non-diabetics [median (IQR); 2.75 (0.8), p = 0.001]. Presence or absence of HbF was comparable between the two groups (p = 0.001). In terms of the presence of HbF (0.3–17.1%), 40.3% had HbF while 59.7% did not have HbF among diabetic patients. In non-diabetic group, 75% had HbF while 25% did not have HbF. All the determined demographic and laboratory characteristics are summarized in Table 1.

Table 1 General demographic and laboratory characteristics of the study participants

Association between HbAIc and age among patients with type 2 diabetes mellitus

The current study considered age as the most important intrinsic demographic that could influence the levels of HbA1c. With diabetics less than 18 years as the reference group, the current study determined association between HbA1c and age in type 2 diabetes mellitus. The study showed that there is a statistically significant association between HbA1c and age [OR = 1.035, 95% CI = 1.007–1.063, p = 0.014] (Table 2).

Table 2 Association between HbAIc and age among patients with type 2 diabetes mellitus

Association between HbA1c and haemoglobin types in type 2 diabetes Mellitus

With non-diabetics as the reference group, the current study determined relationship between HbA1c and haemoglobin types in type 2 diabetes mellitus. Haemoglobin types considered in this study included HbA0, HbA2 and HbF in type 2 diabetes mellitus. The study shows that there is increased association between HbA1c and HbA0 and HbA2 [OR = 1.059, 95% CI = 1.020–0.1.099, p = 0.003] and [OR = 3.893, 95% CI = 2.161–7.014, p = 0.001] respectively. However, HbA1c did not show significant association with HbF [OR = 2.062, 95% CI = 0.873–4.875, p = 0.099] (Table 3).

Table 3 Association between HbA1c and haemoglobin types (HbA0, HbA2, HbSS, and HbF) in type 2 diabetes mellitus

Association between haemoglobin types/ haemoglobin variant and haematological parameters among patients with type 2 diabetes mellitus

With the non-diabetic group as the reference, the correlation between haemoglobin types/variants and haematological parameters was modelled using Spearman’s Correlation. The study showed HCT has a significant correlation with HbA0 and HbAS (r= -0.271, p = 0.021 and r= -0.292, p = 0.013 respectively), while showing a negative correlation with HbF and HbA2 (r= -0.023, p = 0.847 and r= -0.097, p = 0.415 respectively) among the non-diabetics. For MCV, the study revealed a negative correlation with HbAS (r= -0.151, p = 0.206), no correlation with HbA0 and HbA2 (r = 0.198, p = 0.096 and r = 0.223, p = 0.060 respectively), and a significant correlation with HbF (r= -0.291, p = 0.013) in the non-diabetic group. Furthermore, the study showed a significant correlation between MCH and HbF (r = 0.298, p = 0.011), a negative correlation with HbA0 and HbA2 (r= -009, p = 0.938, and r= -0.053, p = 0.660 respectively) and no correlation with HbAS (r = 0.033, p = 0.786) in the non-diabetic group. MCHC had a significant correlation with HbA2 (r=-0.389, p = 0.001), no correlation with HbF and HbAS (r = 0.038, p = 0.748 and r = 0.087, p = 0.468 respectively), while showing a negative correlation with HbA0 (r= -0.068, p = 0.572) for the diabetic group (Table 4).

Table 4 Correlation between red cell parameters and haemoglobin types/variant in patients with type 2 diabetes mellitus

Discussion

Diabetes mellitus is among chronic metabolic disorders that significantly impact on various physiological processes. Type 2 diabetes mellitus is known to act synergistically with other metabolic disruptors leading to life-threatening outcomes, majorly micro-vascular and macro-vascular complications [15]. Diabetes mellitus has previously been described to significantly interfere with multiple hematological parameters [16]. Further, an uncontrolled diabetes status is strongly associated with oxidative stress, inflammation and endothelial dysfunctions that are key players in cardiovascular disorders [17]. When compared to healthy individuals, patients with T2DM have deranged hematological parameters with associated poor health outcomes [18].

The study established a significant association between age and uncontrolled HbA1c in type 2 diabetes mellitus, illustrating a higher tendency of unregulated HbA1c status with advancement in age. These findings are consistent with other studies that observed a higher prevalence and increased risk of complications in older patients compared to younger counterparts [19]. Older population (above 60 years) are approximately 15% of the estimated 7.5 billion people worldwide, out of which 20% have diabetes mellitus while 30% have impaired glucose regulation and are at a higher risk of developing DM [20, 21]. These observations are attributed to various factors, among them long life expectancy that leads to low insulin secretion and resistance, reduced physical activity leading to obesity and other lifestyle-related complications, presence of arginine vasopressin (AVP) that lowers insulin sensitivity, as well as vitamin D deficiency that contributes to osteoporosis and insulin resistance, among others [20].

In the current study, participants with T2DM were found to have lower HbA0-which is a normal proportion of haemoglobin-as compared to non-diabetic participants. Anaemia is a common feature for chronic diabetes mellitus, which is explained by prolonged erythropoietin-deficiency in both type 1 and type 2 DM [22]. This also explains the lower HbA2 levels observed among diabetic participants as compared to the control group, since HbA2 forms part of the normal haemoglobin proportions. On the other hand, a higher proportion of the non-diabetic group had HbF as compared to the diabetic group. This finding corroborates that of a previous study in a South Korean population, which indicated that a long-term glycaemic control together with recent plasma glucose levels do not enhance production of HbF in diabetic patients [23]. The current study attributes this variation in results to glycation of a proportion of HbF among the diabetic patients, which occurs at a slower rate compared to HbA glycation [24]. However, a study by Pardini et al., (1999) found higher levels of HbF among diabetic participants as compared to non-diabetics, concluding that HbF could be an indicator of poor metabolic control in diabetes mellitus. This variation in finding could be due to the heterogeneous nature of the participants in both studies with factors such as age and genetic diversity likely to influence the outcomes.

Using the participants with controlled blood glucose as a reference group, the study also sought to determine a possible relationship between HbA1c and haemoglobin types among the diabetic patients. From the analysis, HbA0 and HbA2 had significant association with HbA1c. However, HbF was not significantly associated with HbA1c. Numerous studies have been carried out to establish the effect of HbF on the measurement of HbA1c, however, there is minimal focus on the relationship between aforementioned Hb types, particularly the association with HbA0 and HbA2. Previously, higher levels of HbF have been observed to significantly interfere with the estimation of HbA1c using various measurement methods [24,25,26]. Thus, this study uniquely establishes the existence of a strong association between HbA1c, HbA0 and HbA2, which should be explored further to determine possible interference in HbA1c estimation, especially among populations with prevalently higher proportions of the highlighted Hb types. Similarly, another study targeting age-specific effects of Hb types on HbA1c interpretation reported a rather low degree of correlation between HbF and Hba1c, suggesting that age could be a significant factor in the interpretation of HbA1c levels [27]. Although the current study majorly focuses on patients of adult age, the findings are consistent with the previous studies indicating a rather low correlation between HbA1c and HbF.

Among other key concerns of the current study was to determine the relationship between haemoglobin types/variants and haematological parameters in the context of type 2 diabetes mellitus. For the diabetic participants, HCT was found to have a significant correlation with both HbA0 and HbAS for the control group while showing a negative correlation with HbAS among the diabetics, although the correlation was not statistically significant. Previously, studies have majorly focussed on the haematological derangements due to hyperglycaemia in diabetes mellitus. This study presents a unique approach by investigating the correlation between individual parameters and the specific haemoglobin types and variants, outlining the possible effects of prolonged hyperglycaemia in T2DM on the highlighted parameters. Although varied observations were made between different parameters and Hb types and variants across both the participant and control groups, individual interpretation of the correlations provide insightful information on the influence of the diabetic status.

Despite there being no significant correlation between a majority of the RBC parameters and the main normal Hb type (HbA0), the study demonstrated the existence of possible interferences of a diabetic status on the interplay between haematological parameters and Hb types/variants. For instance, RBC count, Hb levels, and MCH showed a consistently negative correlation with HbA2 and HbAS while HCT and MCV negatively correlated with HbF and HbAS among the diabetic participants. These findings illustrate the disruption in the formation and composition of haemoglobin orchestrated by chronic insulin resistance (IR) and hyperglycaemia. The observations can be explained by the long-term effects of hyperinsulinaemia on erythropoiesis, which is hypothesised that insulin works synergistically with erythropoietin to influence erythroid progenitors in the process of RBC formation [28]. Furthermore, the increased production of inflammatory markers such as Interleukin-6 (IL-6) and Tissue Necrosis Factor-alpha (TNF-α) as well as an influx of hypoxia-inducible factor-1 alpha (HIF-1 α) due to hyperinsulinaemia are known to disrupt the erythropoietic process leading to IR-related complications in T2DM [29]. Thus, the current study coincides with the existing literature to uniquely describe the interference on haematological parameters and Hb types/variants caused by the diabetic status in T2DM.

Conclusions

HbA1c is associated with increased haemoglobin types (HbA0 and HbA2) but not HbF in type 2 diabetes. Therefore, during routine management of T2DM patients, the levels of various haemoglobin types should be considered in the final HbA1c interpretation. In controlled diabetes HCT had a negative correlation with HbAO and a positive correlation with HbAS while MCV and MCH has a negative correlation with HbF. In uncontrolled diabetes MCH has a negative correlation with HbA2 which should be further considered in the interpretation of HbA1c. The study also concludes that trials focusing on anthropometric and socio-demographic factors are necessary and that targeting representation at community level including undiagnosed population should be studied.

Recommendations

The authors recommend that more comprehensive studies that take into consideration the anthropometric and further socio-demographic factors should be undertaken. Moreover, the authors recommend larger studies using more representative study participants, particularly by targeting representation at community level which may consist of largely undiagnosed population. Consequently, enhancing diagnosis at community level and instituting early management for T2DM is essential in minimizing complications and improving on health outcomes.