The relationship of blood glucose and haemoglobin A1 levels in diabetic subjects

Diagnostic Accuracy of Glycated Haemoglobin and Average Glucose Values in Type 2 Diabetes Mellitus Treated with Premixed Insulin

INTRODUCTION

Studies assessing the relationship between glycated haemoglobin (HbA1c) and average blood glucose (ABG) were conducted in small groups of patients on different treatments and may be biased for these reasons. The aim of the study was to assess the relationship between HbA1c and ABG in a large group of type 2 diabetes patients treated with premix insulin.

METHODS

In 4257 patients treated with premixed insulin, the parallel point-of-care assessment of HbA1c and ABG from the preceding 90 days (ABG90), calculated automatically from all values measured by the glucometer, was performed twice. The regression formulas and respective values of HbA1c and ABG90 were calculated.

RESULTS

The mean number of recorded glucose values/patient was 2.37 estimations per day. The regression formula calculated using data from the first assessment was HbA1c = 5.28 + 0.01487 × ABG90 and that using data from the second one was HbA1c = 4.78 + 0.01683 × ABG90. The slopes of the regression lines are lower than that in a similar analysis from the A1c-Derived Average Glucose (ADAG) study. The comparison of ADAG formula and the formula derived from the present study shows that both formulas give similar results at low HbA1c values, but differ at higher HbA1c values. Additionally, the 95% confidence interval is broader in the PROGENS study e.g. a 95% probability of certainty that the actual HbA1c value was greater than 7.0% (53 mmol/mol) was achieved only at an ABG90 value of 220 mg/dL.

CONCLUSION

The relationship between HbA1c and ABG estimations may be different in various patients; therefore, it seems that the use of one equation in all populations may not be reliable. Broad assessment of ABG as a tool that may replace HbA1c measurements should be recommended only with caution, providing the possible limitations and confidence intervals.

FUNDING

Bioton S.A.

Is there a relationship between mean blood glucose and glycated hemoglobin?

Measurement of hemoglobin A1c (HbA1c) is considered the gold standard for monitoring chronic glycemia of diabetes patients. Hemoglobin A1c indicates an average of blood glucose levels over the past 3 months. Its close association with the risk for the development of long-term complications is well established. However, HbA1c does not inform patients about blood glucose values on a daily basis; therefore, frequent measurements of blood glucose levels are necessary for the day-to-day management of diabetes. Clinicians understand what HbA1c means and how it relates to glucose, but this is not the case with patients. Therefore, the translation of the HbA1c results into something more familiar to patients seemed a necessity. The scope of this article is to review the literature to search for enough scientific evidence to support the idea of a close relationship between HbA1c and mean blood glucose (MBG), and to justify the translation of HbA1c into something that reflects the MBG. Most studies confirm a close relationship between HbA1c and MBG, although different studies result in different linear equations. Factors affecting this relationship may limit the usefulness and applicability of a unique mathematical equation to all diabetes populations.

HbA1c during pregnancy: Its relationship to meal related glycaemia and neonatal birth weight in patients with diabetes

OBJECTIVE

Although home blood glucose (HBG) profiles correlate closely with HbA1c, the strength of the relationship during pregnancy is unclear due to physiological changes which can induce subnormal HbA1c levels. We therefore aimed to establish the strength of the association between mean HBG profiles and HbA1c in diabetic pregnancies and whether HbA1c levels and glycaemic variability affects neonatal birth weight (NBW).

STUDY DESIGN

7-point glycaemic profiles performed throughout pregnancy were obtained retrospectively in 94 consecutive patients attending the diabetes antenatal clinic and compared to the corresponding mean HbA1c levels.

RESULTS

There was a significant linear correlation between mean HBG and HbA1c (HbA1c=0.5HBG+3.1, r=0.71, p<0.0001). Multiple regression analysis demonstrated that both pre- and post-prandial HBG levels correlated significantly and independently with HbA1c, correlation coefficients (r) were 0.63 and 0.65, respectively both p<0.0001. Significant correlations were also observed in patients with gestational diabetes (n=67, mean HbA1c=6.11, r=0.67; p<0.0001) and type 1 diabetes (n=18, mean HbA1c=6.75, r=0.64; p=0.004). All meal related HBG measurements showed similar significant correlations with HbA1c (r values pre- and post-breakfast, pre- and post-lunch, pre- and post-tea and pre-bed are 0.56, 0.55, 0.59, 0.55, 0.56, 0.59, 0.51, respectively p<0.0001 for all time points). Post hoc analysis showed that NBW increased with higher levels of HbA1c; NBW (centiles)+/-S.D. for HbA1c <6.5% versus >6.5% was 78.9%+/-29.2 versus 90.2%+/-18.6, p=0.02.

CONCLUSION

Mean HbA1c levels are closely correlated to all meal related glucose measurements during pregnancy. It is therefore a reliable indicator of overall glycaemic control among patients with diabetes during pregnancy.

Contributions of Weekly Mean Blood Glucose Values to Hemoglobin A1c in Insulin-Treated Type 2 Diabetes: The Diabetes Outcomes in Veterans Study (DOVES)

BACKGROUND

Daily self-monitored blood glucose testing is recommended for patients with insulin-treated type 2 diabetes. However, most patients do not test frequently enough for optimal glycemic control. Less frequent testing may be sufficient for assessing glycemic control among stable patients as well as improving patient compliance. The study objective was to evaluate the weekly contribution of glucose readings to hemoglobin (Hb)A1c during an 8-week period of intensified self-monitored blood glucose testing.

METHODS

The authors randomly selected stable, insulin-treated subjects with type 2 diabetes. Subjects monitored their blood glucose four times daily for 8 weeks; the authors then downloaded glucose meters and measured an HbA1c. Mean blood glucose values were calculated for each of the 8 weeks. Multiple linear regression analyses examined the contribution of these mean values to the HbA1c.

RESULTS

A total of 182 subjects completed the monitoring protocol; mean HbA1c was 7.63 +/- 1.42%, mean glucose was 9.78 +/- 2.27 mmol/L, the regression correlation was 0.77, P < 0.001. A fitted multiple linear model using all 8 weekly mean blood glucose values showed large variation in their independent contributions to the HbA1c. Mean blood glucose values from consecutive weeks were highly correlated and did not provide independent information about glycemic control. Stepwise regression showed that the mean blood glucose values from weeks 4, 6, and 8 significantly and equally influenced HbA1c.

CONCLUSIONS

Glycemic control can be efficiently assessed by reviewing at least 5 weeks' worth of monitoring results, focusing on alternate weeks and giving less weight to more remote readings.

Evaluating once- and twice-daily self-monitored blood glucose testing strategies for stable insulin-treated patients with type 2 diabetes : the diabetes outcomes in veterans study

OBJECTIVE

To evaluate once- and twice-daily self-monitored blood glucose testing strategies in assessing glycemic control and detecting hypoglycemia or hyperglycemia in patients with stable insulin-treated type 2 diabetes.

RESEARCH DESIGN AND METHODS

Subjects with stable insulin-treated type 2 diabetes monitored blood glucose four times daily (prebreakfast, prelunch, predinner, and bedtime) for 8 weeks. We correlated mean blood glucose values with HbA(1c) measured after 8 weeks and determined the number of hypoglycemic (< or =3.33 mmol/l) and hyperglycemic (> or =22.20 mmol/l) readings captured at the various testing times.

RESULTS

A total of 150 subjects completed the monitoring period; their average age was 67 years, 90% were men, and the mean HbA(1c) at baseline was 8.0 +/- 1.8%. The overall correlation of glucose testing and HbA(1c) was 0.79 (P < 0.0001). Mean blood glucose values for each of the four once-daily testing strategies were significantly correlated with HbA(1c) (r = 0.65-0.70, P < 0.0001), as were mean blood glucose values for each of the six twice-daily testing strategies (r = 0.73-0.75, P < 0.0001). The prebreakfast/prelunch measurements captured the largest proportion (63.6%) of the hypoglycemic readings, the predinner/bedtime measurements captured the largest proportion (66.2%) of hyperglycemic readings, and the prelunch/predinner measurements captured the largest proportion (57.7%) of all out-of-range readings.

CONCLUSIONS

Twice-daily testing strategies, particularly prelunch/predinner, effectively assess glycemic control and capture a substantial proportion of out-of-range readings. However, personal testing strategies will vary depending on an individual's risk for hypoglycemia and hyperglycemia.

Monitoring diabetic control in developing countries: a review of glycated haemoglobin and fructosamine assays

Measurement of glycated proteins, especially haemoglobins, is now a routine procedure for monitoring previous glycaemic control in diabetic patients in the developed world. Until recently, however, the assays were sophisticated and costly and, therefore, were only rarely used in developing countries. With easier and cheaper tests now available, the measurement of glycaemic control could be also introduced into regional hospitals in developing countries. This paper reviews feasible methods for clinical laboratories with limited resources. Recent experiences from African countries are included.

Glycaemic control and its determinants in diabetic patients in Ethiopia

A cross-sectional study was undertaken in which concentrations of glycated haemoglobins were measured in 102 diabetics seen at the outpatient clinic in Gondar, Ethiopia, between 26 January and 7 March, 1995. Mean HbA1, levels (standard deviations) were 5.35% (1.1) in non-diabetic controls, 12.0% (1.5) in 59 insulin-dependent diabetics, and 11.0% (2.0) in 43 non-insulin dependent outpatients. The majority of insulin-dependent mellitus (IDDM) (78%) and non-insulin-dependent mellitus (NIDDM) patients (77%) were poorly controlled (HbA1 > 10.8% in IDDM, and > 9.7% in NIDDM, respectively). Multiple linear regression analyses revealed that HbA1 levels were significantly positively associated with lower body mass index, duration of diabetes, a recent history of polydipsia, hypertension, and low income in NIDDM individuals. Whereas in IDDM patients lower age (or alternatively lower age at onset) was the only significant predictor. Whilst 49% of the model variance was explained by the predictors in NIDDM diabetics, only 9% were so in IDDM patients. Current fasting blood glucose level was marginally significant in NIDDM patients (r = 0.29; P = 0.058), but insignificant in IDDM individuals. This points towards the fluctuations in blood glucose levels experienced by IDDM patients in a setting where insulin supply is unreliable. It also confirms the doubts about the usefulness of fasting blood glucose values as a tool for assessing metabolic control.