CLINICAL CONCEPTS
ASSESSING DIABETIC CONTROL
WITH HEMOGLOBIN A1c

The past ten years have witnessed changes in the management of diabetic
patients that promise to banish much of the nihilism formerly
characterizing their prognosis.

For now, new methods providing both long-term and short-term measurements
of carbohydrate in the blood are available.
Long-term glucose measurement can be achieved by testing glycosylated
hemoglobin in peripheral blood. A short-term test of blood glucose can be
obtained at home or work from capillary blood by the patient's using a
reflectance meter, or with visually read strips.

Together, these two new techniques provide both patients and physicians
with an accurate gauge of the glucose status and assist in managing the
diabetes.

Both methods are useful in the management of Type I or IDDM (insulin
dependent diabetes mellitus), formerly referred to as juvenile-onset,
and Type II or NIDDM (noninsulin dependent diabetes mellitus) formerly
referred to as maturity-onset diabetes. But as the nature of these two
type of diabetes differs, so does the emphasis on these two methods.
Type I diabetes implies a lack of insulin and therefore must be treated
with injectable insulin in an attempt to compensate for the drastic
fluctuations of blood glucose levels throughout the day. When a patient
uses a reflectance meter during a given period, he or she can chart
these fluctuations, showing the responses to meals, fasting, and the
insulin dosage. Dosage can then be modified to meet these needs, and
patients can be made aware of their individual responses to diurnal
changes and food intake.

The optimal surveillance frequency for blood glucose measurement is one
that permits documentation of target blood glucose values before and
after meals and allows the prevention of hypoglycemia. We have found
that patients who are initiating intensive insulin delivery and
attempting to normalize hemoglobin A1c values within 6 weeks while
avoiding hypoglycemia, must take approximately 10 measurements of blood
glucose per day. While in practice such aggressive management is not
generally warranted (except in the case of the pregnant diabetic woman),
these observations emphasize the effort required on the part of IDDM
patients to normalize blood glucose safely while avoiding hypoglycemia.
Measuring glycosylated hemoglobin at weekly to monthly intervals, on the
other hand, shows both patient and physician how well the patient has
responded to management over an extended period of time.

TYPE II DIABETES MELLITUS

Type II diabetes implies that, although there is insulin present, the
hormone is not biologically as active as it might be in lowering blood
glucose because of a secretory and/or a metabolic derangment. The
fluctuations in blood glucose in maturity-onset patients tend not to be
as drastic as those in the juvenile-onset patient. Medication, when
necessary, is usually given in the form of an oral hypoglycemic agent,
and marked swings in glucose levels are not a problem. Once the daily
changes in blood glucose for these patients has been established by
self-monitoring of blood glucose, emphasis shifts to the importance of
long-term control, as revealed in periodic monitoring by hemoglobin A1c
(HbA1c) levels.

DAILY MONITORING OF GLUCOSE

In the Type II patient, home use of a reflectance meter or a strip that
can be read directly (several are now commercially available)
establishes the daily variations in blood glucose. Although Type II
patients are not as concerned with the hour-to-hour adjustment of blood
glucose as Type I patients, they are concerned - and this is especially
true of obese patients - with maintaining normolycemia to avoid
complications and maintaining ideal body weight. The obese patient who
initiates a diet will see striking evidence by his or her home glucose
monitoring values whether and how quickly blood glucose can decrease.
Thus patients learn that food plays a major role in their hyperglycemia.
When food is withheld, for example during a 3-day fast, the fall in
blood glucose is dramatic.

Conversely, this process provides striking evidence of how quickly blood
glucose levels rise after overeating. To facilitate behavior
modification alone, monitoring blood glucose before and after eating is
helpful in these patients. Home monitoring is also useful in measuring
the effectiveness of oral hypoglycemic medication. The patient can
readily see that the medication lowers blood glucose levels.
Furthermore, home monitoring can allay fears of hypoglycemia
accompanying the taking of medication. Many patients worry that they
might experience hypoglycemic episodes during the night and are
reassured that such episodes are not occurring, by monitoring of glucose
levels at 2 or 3 AM.

HbA1c BIOSYNTHESIS

HbA1c is one of the minor hemoglobins, but because of its relatively
high concentration in normal persons (3% to 6% of normal hemoglobin), it
is the one most extensively studied. Because circulating erythrocytes
are incapable of initiating protein syntheses, HbA1c is produced as a
post-translational modification of hemoglobin A0 (or adult hemoglobin).
The rate of modification depends on the mean circulating glucose levels
to which the erythrocyte is exposed. The post-synthetic modification of
hemoglobin A to form HbA1c is nearly irreversible, and its rate of
synthesis reflects the glucose environment in which the erythrocyte
circulates. Hemoglobin A1 is a descriptive term which describes all the
fast hemoglobins which includes HbA1c as well as HbA1a and HbA1b.
Because many of these hemoglobins have glucose or phosphorylated
glycolytic intermediates attached, they are referred to as glycosylated
hemoglobins and also reflect average glucose concentrations over time;
however, the measurement is about 50% higher than is the measurement of
HbA1c (or glycosylated hemoglobin) alone.

Reflecting, as it does, a time-averaged "glucose" value, or the mean
blood sugar concentration, for the previous 8 to 10 weeks, the HbA1c
measurements provide a way to judge overall control for the diabetic
patient. Although not widely used now for screening, HbA1c tests may
assume an active role in the initial diagnosis of maturity-onset
diabetes. Peter Bennett, MD, has reported that the HbA1c measurement by
high performance liquid chromatography has achieved a rate of
specificity and sensitivity similar to that of the glucose tolerance
test. On the other hand, it shares similar problems with the glucose
tolerance test, that is, the interpretation of borderline cases. The use
of HbA1c levels in screening, therefore, may obviate the need for the
rather cumbersome glucose tolerance test in a great many patients in the
future, although problems will remain unless standards and references
are established.

However, for helping to establish the proper treatment regimen and
monitoring the long-term glycemic status of the patient with diabetes,
the use of HbA1c measurement provides an accurate index of the degree of
carbohydrate control. If a person with diabetes comes in with a high
level of HbA1c, well above the upper limit of normal values, both
patient and physician know that much more work has to be done to
regulate his or her disease. As the level begins to descend toward the
normal range, they can observe how well a given regimen is affecting the
patient. The rate of rise of HbA1c in response to elevated blood glucose
levels requires a lag time. A new stable plateau is reached in four
weeks. Thus an optimal sampling time for measuring HbA1c might be about
once a month during periods of changing management. Once optimal control
is established, the measurement reaches a new low plateau in eight to
ten weeks. Therefore, the biological off rate is slower than the on
rate. This rate holds as true for persons with Type I disease and
diabetic pregnant women as it does for those with Type II disease.
A single reading of HbA1c cannot tell the clinician whether the control
is in an ascending or a descending curve. Two measurements are necessary
to establish the direction in which the average blood glucose is
heading, although they are not sufficient to establish what the plateau
is going to be. While the measurement reflects an average blood glucose,
rather than peaks and valleys, for clinical purposes it is certain that
any elevation of blood glucose that has continued for a week or more
will be clearly reflected in a subsequent measurement of HbA1c, assuming
the assay is well performed.

Proper scheduling of the measurements depends on the information the
clinician is seeking. If he or she is looking for the results of the
therapeutic intervention, he should be able to find a difference in the
levels of HbA1c after one week. If a value is sought for a patient whose
therapy has not been changed, then monthly assessments are more than
sufficient. In a mature stable patient who has offered no reason for
unusual concern, the HbA1c need not be measured more often than once
every three to six months.