Diabetes mellitus is a chronic disease, with an increasing prevalence worldwide, Nigeria  inclusive  (Nyenwe  et  al., 2003).  Prevalence in Nigeria has been reported to range from 0.85% in rural Mangu village, to as high  as  11%  in urban Lagos (Akinkugbe, 1967).  The WHO suggested that Nigeria has the greatest number of people living with diabetes in Africa, with an estimated burden of about 1.7 million, which will increase to 4.6 million by 2030 (Chineye et al., 2012). Diabetes Mellitus leads to both microvascular and macrovascular complications, which include cardiovascular and renal complications if not treated early. Prevention of the disease is of utmost importance. Early diagnosis and treatment, maintaining good glycaemic control helps to reduce co-morbidity and mortality. Glycated haemoglobin is not only a diagnostic tool for diabetes mellitus; it is also a measure of glycaemic control over the past 2-3 months in an individual (Sacks, 2007). In patients with diabetes mellitus, glycated haemoglobin value is used to determine the degree of glycaemic control and make decisions regarding therapy. In addition the concentration of glycated haemoglobin predicts the progression of diabetic microvascular complication (Sacks, 2007; UK Prospective diabetes study (UK PDS, 1998). More recent evidence indicates that glycated haemoglobin also predicts cardiovascular disease in patients with Type 1 diabetes (Sacks, 2007; American Diabetes Association, 2007). Measurement of glycated haemoglobin is thus an important component in the management of patients with diabetes (American Diabetes Association, 2016). The American Diabetes association recommends that HbAIc should be measured at least twice a year (American Diabetes Association, 2018) and Good glycemic control is reported at HbAIc at <7% (American Diabetes Association, 2018).

The relationship between glycated haemoglobin (HbA1c) and Estimated Average glucose (eAG) is described by the formula; eAG - 28.7 × AlC - 46.7 (Cha and Ko, 2016; Raja et al., 2003), which is Nathan’s regression equation recommended by the American Diabetes Association (Raja et al., 2003).

Cha and Ko (2016) have said translating the A1c value into an estimated average glucose (eAG) is more practical and much easier for patients to understand.  Riyadh and Abdullah, (2018) and Zhou et al. (2013) have reported glycaemic control findings in their area of study. The need to assess glycaemic control and determine the correlation between glyceated hemoglobin and estimated average glucose has become important due to the increasing prevalence of diabetes milletus in Nigeria.  

Data of l00 diabetic patients (38 males and 62 females) seen consecutively in the endocrinology outpatient clinic between March 2017 and October 2017 were analyzed. They were patients already diagnosed with diabetes mellitus according to WHO criteria and were on blood glucose lowering drugs. Relevant data of both Type 1 and 2 Diabetic patients whose blood samples were sent for glycated haemoglobin assay, were retrospectively analyzed.  Other demographic data such as age, gender and biochemical investigations of the patients were retrieved and analyzed.  Hospital ethical committee protocol was observed. Glycated haemoglobin (HbA1c) assay was done using high performance liquid chromatography.  (D₁₀ – Bio Rad). Estimated average glucose was calculated using Nathan’s regression formula;

eAG = 28.7 × HbAlC – 46.7

Statistical analysis

This was done using Statistical Package of Social Sciences (SPSS) version 21. Difference between means of variables was determined using the student “t” rest.  Correlation between variables was examined using Pearson’s correlation. Level of significance was set at P≤ 0.05.

Data of a total of 100 diabetic subjects (38 males, 62 females) were analyzed. Mean age of the subjects was 61.5 ± 14.8 years (Table 1). Mean fasting plasma glucose, estimated average glucose (eAG) and glycated haemoglobin in the diabetic subjects were 189.6 ± 100.3, 165.4 ± 74.8 mg/dl, 7.4 ±2.4% respectively (Table 1). Mean fasting plasma glucose, estimated average glucose and glycated haemogobin were 183.6 ± 93.0. 158.1±85.3 mg/dl, 7.3 ± 2.8%, respectively in the male subjects and 196.1 ± 112.0, 169.9 ± 68.3 mg/dl, 7.5 ±2.2% in the females (Table 2). Difference was not statistically significant (P > 0.05). Estimated average glucose showed a strong positive correlation with glycated haemoglobin r =1.000   P = 0.000. (Table 3 and Figure 1). 46% of the subjects had glycated haemogobin <7% while 39% had levels between ≥ 7 and 10%;  15% had levels > 10% (Table 4).

With the increasing prevalence of diabetes mellitus, the social and economic burden on the nation is rising. Adequate prevention strategies and early diagnosis, intervention and treatment to ensure good glycaemic control are of utmost importance.  Mean age of the subjects in the study was similar to that reported by Shittu et al. (2017). There were more females in this study than males. Similar findings were reported by Shittu et al. (2017) in their study in Oke Osun Region of Oyo State, Nigeria and the Nigerian National Communicable (The Expert Committee on Non-communicable Disease (1997) diseases. Other studies (Elmandi et al., 1991) also made similar observations. The combined effect of more elderly women than men in most populations is the most likely reason for this observation (Shittu et al., 2017). However, some authors have reported male preponderance (Amooh et al. (2002).

Mean fasting plasma glucose, glycated haemoglobin levels were elevated in both the male and female subjects. Though,  there  were  no  statistically  significant differences between the values in the male and female subjects. Estimated average glucose showed a positive correlation value of 1 with Glycated haemoglobin in our study. Several authors have reported similar findings. Nepal et al.¹⁷ reported a strong correlation value of 1, which was statistically significant, in their study, and concluded that “blood glucose expressed as eAG improves the understanding of blood glucose monitoring”. Nathan ¹⁸ et al. recorded a correlation value of 0.8, which was statistically significant and concluded that glycated haemoglobin  levels  can  be   expressed   as    estimated average blood glucose for most patients with type 1 and type 2 diabetes. Sacks (2007) and American Diabetes Association (2016) also reported linear correlation of estimated average glucose and glycated haemoglobin. From these findings, reporting Glycated haemoglobin values along with estimated average glucose levels, will definitely be beneficial in the management of Diabetes Mellitus patients as it plays a dual role of giving the physician a quick overview of average blood glucose over the preceding 2 to 3 months and helps to determine the type    of  pharmacotherapy  and  necessary  adjustments where applicable. Also, it is an easier and simpler interpretation to the patient on how his/her blood glucose has been controlled over the past 2-3 months. This may also boost compliance, especially if the estimated average blood glucose is still yet to be within the normal range, the patient knows he/she would have to take his/her medications. From the finding in our study, reporting estimated average glucose along with glycated haemoglobin may be beneficial in monitoring and assessment of long term glycaemic control in diabetics in our area of study. The American Diabetes Association and the American Association of Clinical Chemist have determined that the correlation (r=0.92) in the ADAG trial is strong enough to justify reporting both the A1c results and the estimated average glucose result when a clinician orders the A1c test¹⁰.  However, fasting plasma glucose estimation is still the tool for short term monitoring of glycaemic control and is a relevant part of diabetic management protocol. Reporting the three values together gives a holistic picture of both the past and present glycaemic status of the patient.

There are different schools of thoughts on good glycaemic control. While the American Diabetes Association recommends good glycemic control to be HbA1c > 7%, International Diabetes Federation and American College of Endocrinology recommends values < 6.5% as good glycaemic control. HbA1c < 7% was considered as good glycaemic control in our study. Only 46% of the subjects in this study had glycated haemoglobin < 7 and 54% had glycated haemoglobin values  ≥7%,  with  15%  showing  poor  glycemic  control having values > 10%. Idogun and Olumese recorded 46.8% with poor glycaemic control in 2007, Unadike et al. (2010) reported 46% in 2010, Edo E (Edo and Akhiemokhan, 2012) reported 63% with poor glycaemic control in 2012, also in a previous study by Adewolu, (2014), 59.5% had poor glycaemic control, all in the same area of study. Adebisi et al. (2009) and John et al. (2005) in Nigeria have also reported high prevalence of poor glycaemic control. Achieving good glycaemic control of diabetic patients is still a challenge. Factors ranging from non-compliance to pharmacotherapy, lack of funds to purchase required medications and investigations, payment of hospital bills, socio-cultural beliefs may all be contributory to this development. Economic reasons may take a larger percentage of these factors because of the prevailing high cost of living. Measures such as good patient education, drug subsidization, expansion of scope of health insurance schemes to include diabetes care may be helpful.

In this study, estimated average glucose correlated strongly and significantly with glycated haemoglobin in the diabetic subjects, therefore reporting estimated average glucose along with average haemoglobin will be useful and beneficial in monitoring of long term glycaemic control. 54% of the subjects had poor glycaemic control.

Reporting estimated average glucose (eAG) along with glycated haemoglobin may be useful and beneficial in the assessment and monitoring of long term glycaemic control in diabetic patients. Glycaemic control is yet to be optimal in the study population . Socio-economic measures and expanded health insurance coverage may be useful in optimizing glycaemic control.

The authors have not declared any conflict of interests.