Glycosylated haemoglobin and steady-state mean blood glucose concentration in type 1 (insulin-dependent) diabetes

Diabetologia, Nov 1982

Summary Since glucose control and glycosylated haemoglobin varies asyncroneously, we have studied the steady-state relationship between these two factors. In Type 1 (insulin-dependent) diabetic patients with a constant haemoglobin A1c during the preceding 2 years, 15 ambulatory blood glucose profiles during a 5-week period showed a constant glucose level and provided a precise estimate of the mean blood glucose concentration. In addition, we studied 15 non-diabetic subjects who provided three glucose profiles and had one haemoglobin A1c determination performed. A good correlation was found for a curvilinear relationship (haemoglobin A1c=2.07 x mean blood glucose0.596, r=0.98). This close relationship indicates that glycosylated haemoglobin is a valuable, but not very sensitive, index of glucose control.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://link.springer.com/content/pdf/10.1007%2FBF00260951.pdf

Glycosylated haemoglobin and steady-state mean blood glucose concentration in type 1 (insulin-dependent) diabetes

Diabetologia Glycosylated Haemoglobin a n d Steady-State Mean Blood Glucose Concentration in Type 1 (Insulin-Dependent) Diabetes P. A a b y Svendsen T. Lauritzen U. S o e g a a r d a n d J. N e r u p Summary. Since glucose control and glycosylated haemoglobin varies asyncroneously, we have studied the steady-state relationship between these two factors. In Type 1 (insulin-dependent) diabetic patients with a constant haemoglobin AI~ during the preceding 2 years, 15 ambulatory blood glucose profiles during a 5-week period showed a constant glucose level and provided a precise estimate of the mean blood glucose concentration. In addition, we studied 15 non-diabetic subjects who provided three glucose profiles and had one hae0 0 1 2 - 1 8 6 X / 8 2 / 0 0 2 3 / 0 4 0 3 / $ 0 1 . 0 0 Long-term mean blood glucose concentration; glycaemic control; haemoglobin A~c; Type 1 diabetes - 9 Springer-Verlag 1982 I n the m a j o r i t y o f insulin-treated T y p e 1 diabetic patients, b l o o d glucose concentrations fluctuate considerably, m a k i n g precise a s s e s s m e n t o f l o n g - t e r m g l y c a e m i c control in the average o u t p a t i e n t very difficult to obtain. F o r this reason, l o n g - t e r m g l y c a e m i c control m i g h t be m o r e precisely reflected b y the glycosylated fraction o f h a e m o g l o b i n . Several reports h a v e s h o w n significant correlations b e t w e e n glycosylated h a e m o g l o b i n a n d exp r e s s i o n o f m e t a b o l i c control in cross-sectional as well as longitudinal studies [ 1-7 ]. M o s t studies r e p o r t e d r-values b e l o w 0.9 i m p l y i n g that a given c o n c e n t r a t i o n o f glycosylated h a e m o g l o b i n m a y result f r o m varying degrees o f glucose control. Obviously, this m a k e s the usefulness o f glycosylated h a e m o g l o b i n d e t e r m i n a t i o n in the a s s e s s m e n t o f l o n g - t e r m g l y c a e m i c control in the individual patient r a t h e r doubtful. Since stable h a e m o g l o b i n Ale (HbAI~) r e s p o n d s slowly to changes in glucose control [ 8, 9 ], it m i g h t be t h a t the real c o n n e c t i o n b e t w e e n integrated glucose control a n d H b A l c in T y p e 1 diabetic subjects can be established only in patients k n o w n to be in stable glucose control o v e r a long p e r i o d o f time. T o test this possibility, we h a v e studied the correlation b e t w e e n m e a n b l o o d glucose c o n c e n t r a t i o n a n d stable HbA1~ in diabetic patients with stable m e t a b o l i c control for m a n y weeks. moglobin Ale determination performed. A good correlation was found for a curvilinear relationship (haemoglobin Ale = 2.07 mean blood glucose~ r = 0.98). This close relationship indicates that glycosylated haemoglobin is a valuable, but not very sensitive, index of glucose control. Subjects o f 38 randomly selected insulin-treated diabetic outpatients (age: 15-81 years, duration of diabetes: 3-37 years) in whom several HbAlc determinations had been performed over a 2-year period, 15 patients had six or more HbAlcdeterminations showing a coefficient of variation of < 10% (mean 5.2%).Detailed information about the patients is given in Table t. These 15 patients consented to collect 15 profiles of seven samples each for blood glucose determination over a 5-week period. HbAlo was determined again immediately after the study period and this value was used in the correlation analysis. Fifteen non-diabetic volunteers (eight women, seven men) from the hospital staff served as control subjects and provided three seven-sample profiles and had one HbAlcdetermination performed. Methods At home patients and control subjects collected seven 10 lxl blood samples per day by finger-pricking, using insulin-injection needles and 10 ~tl end-to-end capillaries. Samples were obtained 30 rain before and 90 rain after the three main meals and at bedtime. Reagents of the Glucoquant-Haemolysate system (Boehringer, Mannheim, FGR) were used to stabilize glucose during mailing by ordinary mail to our institute and for determination of glucose in the hospital laboratory. The results of glucose determinations obtained by the Glucoquant system were compared with the routine laboratory method [ 10 ] in two ways, (1) single samples taken by 41 patients were compared 5 1'0 1'5 2'0 2'5 Mean Blood glucose (mmol/I) with the routine laboratory method on blood taken simultaneously by laboratory technicians and (2) mean blood glucose values obtained by the two methods of blood sampling and analysis were compared on the basis of six-sample profiles in 28 patients. Daily mean blood glucose values were calculated as the arithmetic mean of all samples a day and the overall mean blood glucose and its variation for each patient and control calculated from daily mean blood glucose values. To ensure that seven-sample profiles accurately reflect the 24-h mean blood glucose values, mean blood glucose values from sevensamples were compared with mean blood glucose values calculated from 31-sample profiles in 34 Type 1 diabetic patients, where both sets of samples were obtained. Stable HbAac was determined as described previously [ 11 ] after incubation of blood samples diluted fivefold in saline (0.15 tool/l) for 6 h at 37~ C. A fresh blood sample from one non-diabetic person was determined in all analytical series and showed a day-to-day variation of 3.6% (coefficient of variation) during the study period. HbAac results are expressed as percentage of total haemoglobin. Linear and curvilinear regression analyses were performed using standard programmes of the HP-85 calculator. Results are expressed as mean _+ SD unless otherwise specified and differences in mean values are tested by the Mann-Witney ranksum test. Comparison between glucose values from samples taken by patients (Glucoquant Haemolysate system) P.A. Svendsen et al.: HbAac and Steady-State Mean Blood Glucose Concentrations Diabetes duration (years) Mean 8.6 Range and laboratory method demonstrated good agreement between the two methods. The 41 single samples gave values only 0.01 _+ 0.71 mmol/1 lower in the Glucoquant system and also the 28 single day, six-sample mean blood glucose values differed only slightly (on average 0.1 + 0.5 mmol/1). These differences were considered unimportant. Mean blood glucose values calculated from seven-sample profiles in 34 Type 1 diabetic patients were slightly higher than mean blood glucose values calculated from 31-sample profiles taken on the same day (the mean difference being 0.4 + 1.1 mmol/1). This difference was considered acceptable. The mean blood glucose values of the 15 insulintreated out-patients ranged from 6.5-22.0 mmol/1 with standard deviations ranging from 0.84.3 mmol/1. The difference between the mean of all glucose values from the first and last 2 weeks of the study period was significant in three patients only. No difference was found when daily mean blood glucose values were compared for the two periods and therefore we considered most patients to have a constant glucose level. In the 15 control subjects, the standard deviations of mean blood glucose determinations ranged from 0.03 to 0.5 mmol/l. HbA lc Determinations The HbAtc values for the 15 patients ranged from 6.2-12.8%. The control subjects had values ranging from 4.7-5.8%, i.e. within the normal range (4.1-6.4%) for our laboratory. The Mean Blood Glucose-HbAlc Relationship Figure 1 shows the results of mean blood glucose- and HbAlc-values plotted against each other. For the combined material of Type 1 diabetic and control subjects, the coefficient of correlation was highest for the curvilinear correlation (HbAoo = 2.07 x mean blood glucose0'596, r -----0.98). Discussion In most insulin-treated Type 1 diabetic patients, blood glucose concentrations vary greatly within as well as between days. Long-term assessment of diabetic control in Type 1 diabetic patients based upon glucose determinations would require large numbers of 24-h profiles produced by large numbers of blood glucose determinaP. A. Svendsen et al.: HbA1oand Steady-State Mean Blood Glucose Concentrations tions, thus being difficult to obtain, expensive and rather d e m a n d i n g for the patients. Since HbA~ seems to reflect integrated blood glucose control over long periods of time [ 12 ], we f o u n d it o f interest to investigate whether the varying correlations between different parameters of glucose control and HbAt fractions reported so far [ 1-7 ] might be due to lack of precision of the m e a n blood glucose estimates in the typical insulin-treated diabetic patient. By studying patients expected to be in relatively stable blood glucose control and by frequent sampling of seven sample blood glucose profiles, we were able to demonstrate a very strong and curvilinear correlation between m e a n blood glucose and stable HbAtc (r = 0.98). A slightly curved relationship between m e a n blood glucose and HbAtc is predicted by the mathematical model proposed by Beach [ 13 ] and has been p r o p o s e d by others [ 5-7 ] on the basis of clinical observations. However, most of the coefficients of correlation reported so far were too low to allow the use of HbAIo determination as a precise tool in the assessment of m e a n b l o o d glucose or long-term glycaemic control in the individual out-patient Type 1 diabetic. The biological significance of the curvilinear relationship is unknown, but it does make HbAtc a somewhat insensitive index of integrated blood glucose control in the higher range of m e a n blood glucose levels. The obvious practical consequence of the curved relationship is that it is necessary to introduce HbAtc assays with high analytical specificity and precision. Our results demonstrate that m e a n blood glucose m a y be precisely estimated from one H b A ~ determination in insulin-treated diabetic patients with a constant m e a n blood glucose level. In ongoing studies, we have observed that patients in w h o m m e a n blood glucose values are calculated from less frequently sampled profiles and selected on the basis o f ( l ) constant subcutaneous insulin infusion, (2) retinopathy, (3) nepropathy, and (4) HbAlc values above 10%, also scattered a r o u n d the regression curve f o u n d in this study. Acknowledgements. We thank Mrs. A.Rafn for excellent secretarial assistance. 1. Gonen B , Rochman H , Rubenstein AH , Tanega S ( 1977 ) Haemoglobin A1: an indicator of the metabolic control of diabetic patients . Lancet II: 734 - 736 2. Berger W , Sonnenberg GE ( 1980 ) Blutzuckertagesprofile and H~moglobin A1 zur Oberwachung der Diabetesbehandlung . Schweiz Med Wochenschr l l 0: 485 ~ 91 3. Paisey RB , MacFarlane DG , Sherriff RJ , Hartog M , Slade RR , White DA ( 1980 ) The relationship between blood glycosylated haemoglobin and home capillary blood glucose levels in diabetics . Diabetologia 19 : 31 - 34 4. Daubresse JC , Lemy C , Bailly A , Meunier JC ( 1979 ) The usefulness of a rapid method for total fast haemoglobin determination in screening for diabetes control . Diabete Metab: 5 : 301 - 305 5. Graf RJ , Halter JB , Porte D ( 1978 ) Glycosylated haemoglobin in normal subjects and subjects with maturity onset diabetes . Diabetes 27 : 834 - 839 6. Topper E , Doron M , Karp M , Laron Z ( 1979 ) Relationship between glycosylated haemoglobin and postprandial blood glucose in insulin-dependentjuvenile diabetes patients . Isr J Med Sci 15 : 283 - 284 7. Stanton KG , Davis RE ( 1978 ) The relationship between the control of diabetes mellitus and circulation glycosylated haemoglobin A1 . Aust NZ J Med 8 : 400 - 404 8. Koenig RJ , Peterson CM , Jones RL , Saudek C , Lehrman M , Cerami A ( 1976 ) Correlation of glucose regulation and haemoglobin Ale in diabetes mellitus . N Engl J Med 295 : 417 - 420 9. Peterson CM , Jones RL , Dupuis A , Levine BS , Bernstein R , O'Shea M ( 1979 ) Feasibility of improved glucose control in patients with insulin-dependentdiabetes mellitus . Diabetes Care 2 : 329 - 335 10. Trinder P ( 1969 ) Determination of blood glucose using glucose oxidase with an alternative oxygen acceptor . Ann Clin Biochem 6 : 24 - 27 11. Svendsen PA , ChristiansenJS, Soegaard U , Welinder BS , Nerup J ( 1980 ) Rapid changes in chromatographically determined haemoglobin A~ induced by short-term changes in glucose concentration . Diabetologia 19 : 130 - 135 12. Dunn PJ , Cole RA , Soeldner JS , Gleason RE ( 1981 ) Stability of haemoglobin Ale levels on repetitive determination in diabetic out-patients . J Clin Endocrinol Metab 52 : 1019 - 1022 13. Beach KW ( 1979 ) A theoretical model to predict the behaviour of glycosylated haemoglobin levels . J Theor Biol 81 : 54 %561 Received: 2 September 1981 and in revised form: 16 June 1982


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2FBF00260951.pdf

P. Aaby Svendsen, T. Lauritzen, U. Søegaard, J. Nerup. Glycosylated haemoglobin and steady-state mean blood glucose concentration in type 1 (insulin-dependent) diabetes, Diabetologia, 1982, 403-405, DOI: 10.1007/BF00260951