International Health News

How Do You Know If You Have Diabetes or Pre-Diabetes?

by William R. Ware, PhD

Bill Ware

"How Do You Know If You Have Diabetes or Pre-Diabetes?" is an important question given the current epidemic of adult-onset diabetes. The short answer is - ask your doctor who will order a test. However, as we shall see, it is far from that simple. Another short answer is that it simply a matter of a definition, and that there are more than one.

Diabetes is a metabolic disease characterized by abnormally elevated levels of blood glucose (hyperglycemia) resulting from defects in insulin secretion, insulin action or both. Chronic hyperglycaemia of diabetes in associated with the dysfunction, long-term damage and failure of various organs and systems, especially the kidneys, eyes, nerves, heart and blood vessels. These long-term complications include retinopathy with potential loss of vision, nephropathy leading to kidney failure, peripheral neuropathy with risk of foot ulcers, amputations, and neuropathy of the autonomic nervous system resulting in gastrointestinal, genitourinary and cardiovascular symptoms and sexual dysfunction. Thus diagnosing diabetes or pre-diabetes and then attempting to reverse the associated hyperglycemia or at least minimize the micro- and macrovascular damage is an important and significant challenge with profound public health implications.

The vast majority of cases of diabetics are so-called Type 2, which historically was termed adult-onset diabetes and is caused by a combination of resistance to the action of insulin and an inadequate compensatory insulin secretory response. Type 1 diabetes, on the other hand, is caused by an absolute deficiency of insulin secretion, frequently occurs at a young age and requires generally requires life-long insulin injections. In Type 2 diabetes the degree of hyperglycemia sufficient to cause pathologic and functional changes in target tissues may be present for long periods of time before diabetes is detected, although observable abnormalities in carbohydrate (glucose) metabolism are present. Thus in order to prevent vascular problems, it is very desirable to detect not only the presence of diabetes but to identify and treat the pre-diabetic state. This review is concerned with Type 2 diabetes.

It is estimated that 30% of individuals with diabetes are undiagnosed and of these, 25% already have microvascular complications at diagnosis, suggesting the disease had pre-existed for more than 5 years. One commonly used criterion for the presence of diabetes is elevated fasting glucose. However, it has been found that for the vast majority of patients, only a non-fasting glucose is measured - the so-called random or casual value. One study found only 3% of screening for diabetes was done with fasting plasma glucose.[1] This is surprising since measuring cholesterol requires fasting, and practically everyone seems to know their cholesterol status. The random glucose approach is a matter of convenience, but the result is strongly dependent on when the sample is drawn relative to the last meal and as well as what was eaten. There is no agreement on what random glucose value should lead to further investigation into the question of the presence or absence of diabetes or pre-diabetes, and the non-fasting glucose level used in the diagnostic guidelines is supported by very little evidence.

My favourite Supplements

The progression to diabetes is characterized by a continuum of changing characteristics of glucose metabolism which are reflected in both fasting glucose and the magnitude and time evolution of the blood glucose response to food intake. Thus fasting glucose is just one of the indicators of the state of glucose metabolism. Another traditional tool for observing abnormalities in glucose metabolism is the oral glucose tolerance test (OGTT) which measures the blood glucose level over time after drinking a solution of glucose to end a fast. The blood glucose level at 2 hours is diagnostic for both diabetes and insulin resistance. For the diagnosis of diabetes, the approach has been to select glucose levels that reflect thresholds for microvascular-associated complications of hyperglycemia. Thresholds in this context are quite fuzzy and thus diabetes diagnostic criteria or evidence of impaired glucose metabolism are somewhat arbitrary. The tendency appears to be to err on the conservative side given that undiagnosed diabetes presents serious health risks which may justify tolerating a certain level of false positive results.

The most recent (2008) guidelines from the American Diabetes Association (ADA)[2] and the World Health Organization (WHO, 2003)[3] are almost identical. Traditionally, the OGTT was the "gold standard" for diagnosing diabetes. Now there is a choice, i.e. two "gold standards! In what follows the units for blood glucose used will be mg/dL and the alternative unit used in Europe and Canada, the mmol/L, will be given in parentheses. The conversion factor is 18 mg/dL per 1 mmol/L.

  • Fasting plasma glucose (FPG) equal to greater than 126 mg/dL (7.0) OR
  • A plasma glucose level of equal to greater than 200 mg/dL (11.1) at 2 hours after ending fasting with a 75-gram oral glucose challenge, the OGTT.

The guidelines differ only in that the ADA also includes a casual or random plasma glucose level of equal to greater than 200 mg/dL (11.1) and clinical symptoms of hyperglycemia (elevated blood sugar), which cannot be quantitatively specified. Problems with the random glucose test have been discussed above. The usual practice is to confirm a positive diagnosis with a repeat of the same measurement on another day. Thus the situation is simple. There is essentially a choice of two methods, both requiring fasting. Aside from the requirement of fasting, one option is convenient, the other viewed as inconvenient for the patient (sit and wait for 2 hours or return after 2 hours to the office or diagnostic laboratory) and can even induce nausea and vomiting. Because the OGTT is no longer required by the guidelines for positive diagnosis of diabetes, in many practices it has all but disappeared. But it can still be found included in very comprehensive physical exams and as well is used by specialists. But since these two high profile organizations offer an unqualified choice, one might assume that the two tests would essentially identify the same populations, diabetics and non-diabetics. The concern is undiagnosed diabetics, and this implies that either choice in diagnostic approach will reveal who they are. Surprisingly, the data indicate that this is not true.

The results of a large study illustrate the problem. Data were collected from 13 populations in eight European countries and comprised over 17,000 men and 8300 women, age range 17-92. The relevant comparison in the context of this review is the agreement of classification of patients using the two above criteria in a cohort of individuals without a diagnosis for diabetes. In what follows, the term WHO criterion will refer to the diagnosis by an OGTT with a 2-hour post challenge plasma glucose equal to greater than 200 mg/dL since this was the earlier WHO required criterion, whereas the ADA criterion will refer to a FPG equal to greater than 126 mg/dL, the new gold standard. A total of 1517 individuals had diabetes according to either the ADA or WHO criteria. But among the 904 who had diabetes according to the WHO criterion, 473 (52%) had a fasting plasma glucose of < 126 mg/dL mg and thus according to the ADA criterion did not have diabetes. When the same comparison was made for the 1044 who had diabetes according to ADA criterion, 59% failed to qualify for a diagnosis of diabetes according to the WHO OGTT criterion. Thus out of the total of 1517 individuals diagnosed by one or the other or both criteria, there was concordance between the two only for 413 individuals (27%). Individuals below the age of 65 were more likely to be diagnosed on the basis of the ADA criterion, whereas the WHO criterion was more likely to diagnose diabetes in lean individual and the ADA criterion more likely identify those middle-aged and obese[4].

A smaller study published in 2003 reached essentially the same conclusion regarding the lack of concordance of the two criteria. Two hundred and twenty eight patients with a fasting glucose between 110 and 125 mg/dL, i.e. not diabetic according the ADA criterion, were given the OGTT. Using the WHO criterion, 33.6% had diabetes[5]. Other studies have found a similar disparity between using the OGTT as compared to FPG[6].

Fasting glucose also varies from day to day by 12-15% and the laboratory variability is 4%. Using a variability of 15%, a value just below the diabetes diagnostic threshold of 125 could in fact be 144 or 109 mg/dL. There is another problem with the use of fasting plasma glucose as a single criterion for the diagnosis of diabetes. It turns out that the fasting glucose level varies systematically during the day. Individuals with an afternoon appointment who have been instructed to fast starting the night before will on average have a lower fasting glucose level than those with a morning appointment. One study found that the among those tested in the afternoon, difference was such that half of all cases of undiagnosed disease would be missed[7].

It appears that while both the WHO and ADA criteria are equivalent in predicting microvascular diseases such as retinopathy, the OGTT is better at identifying those individuals with increased risk of cardiovascular disease associated with hyperglycemia.6 However, it appears that both must be used to avoid missing a substantial number of individuals who should be tagged "diabetic." Of course, if an OGTT were carried out, a FPG would normally be obtained just before the glucose challenge. Such an approach would considerably increase the number diagnosed, would require the unpopular 2-hour OGTT and presumably because of issues of "convenience" would not be acceptable today for screening except perhaps in the case of high-risk patients (or executives). Thus the failure to diagnose all diabetes cases seems to have been built into the system. An alternative approach, which has been suggested, involves a marker of blood glucose averaged over several months, glycated haemoglobin called HbA1c.

Another way of assessing hyperglycemia is to examine the long-time blood glucose average provided by what is called glycated haemoglobin, which is short for haemoglobin that has reacted with glucose. It is expressed as a percentage of total haemoglobin. Because the reaction of glucose with haemoglobin (glycation) occurs over the entire lifespan of the red blood cells (2-3 months), it provides a long-term measure of the average glucose level. An important feature of this haemoglobin fraction is that fasting is not required for a meaningful determination. Also, testing for HbA1c is already well established and routinely done in connection with monitoring glycemic control among diabetics. Given that HbA1c is in a sense a built-in glucose monitor that can be read with a blood test, it is not surprising that the level of this haemoglobin fraction can be used to identify individuals with diabetes or those at increased risk, and that there is growing interest in this application. What is surprising is that the latest American Diabetes Association (ADA) guidelines still recommend against using this marker[2]

The evidence that HbA1c correlates with the average glucose level is quite strong. To examine the extent of the correlation, it is necessary to establish average blood glucose levels. During the day, the level varies due to post-meal elevations which then decline toward pre-meal values with rates that depend on the presence or absence of diabetes or pre-diabetes. During the night, the levels are typically low and for diabetics frequently increase in the early morning. Thus the ideal approach is interstitial continuous monitoring. In a recent study of a group of individuals which ranged from normal to diabetic, Nathan et al [8] used this continuous monitoring coupled with calibration from eight finger-stick measurements during the day and night. Glucose measurements were automatically made every 5 minutes for at least 2 days at baseline and this was repeated every 4 weeks during the next 12 weeks. In addition, finger stick measurements without the 3:00 A.M. measurement were used for at least 3 days per week when the continuous monitor was not being used. The average glucose level was calculated for each individual from approximately 2700 values collected over 3 months. A linear relationship was observed (AvPG = 28.7 X A1c - 46.7 mg/dL) with excellent correlation (a correlation coefficient of 0.84). An HbA1c value of 5% corresponded to an average glucose level of 97, 6% to 126, 7% to 154 and 8% to 183 mg/dL. Other studies also found linear relationships,[9,10] but the values found by Nathan et al are somewhat lower. The authors regard this as due to the use of continuous monitoring which provided more measurements during the night. These studies appear to quantitatively validate the belief that HbA1c is a reliable measure of average blood glucose levels over several months.

The next issue involves the ability of HgA1c to identify individuals with diabetes. From what was discussed above it is evident that there immediately arises the problem of a diagnostic reference or benchmark. There is no single gold standard for the diagnosis of diabetes. Thus some studies use the traditional WHO OGTT threshold and some use the fasting plasma glucose (FPG) cut-off of = 126 mg/dL. Involved are issues of screening, and thus the sensitivity and specificity. Sensitivity is the percentage of tested individuals at or above a cut-off point who have diabetes. Specificity is the percentage of tested individuals below the cut-off point who do not have diabetes.

Since the latest guidelines essentially make fasting plasma glucose (FPG) equal to greater than 126 mg/dL the criterion for a diagnosis of diabetes, it is of interest to examine heat-to-head comparisons of FPG and HbA1c where the old gold standard, the OGTT, is used as a reference. In a paper published in 2007, Bennett et al [11] reviewed a number of studies in order to examine this question. HbA1c and FPG were similar in their ability to identify diabetics. They found that an optimum cut-off point for HbA1c was > 6.1% which gave a sensitivity of 78-81%, i.e. the percentage of individuals with diabetes who were correctly identified based on the OGTT standard. It should be recalled that when FPG was compared to OGTT in the studies described above, FPG failed to identify approximately 50% of those diagnosed using the OGTT as a reference. The authors also point out that HbA1c exhibits less intra-individual variations and better predicts both micro- and macrovascular complications.

Recent studies have used the FPG cut-off as a reference and examined the predictive power of HbA1c. Ginde et al.[12] examined 6700 individuals undiagnosed for diabetes and representative of the U.S. population for whom HbA1c and FPG were known. FPG equal to greater than 126 mg/dL was taken as diagnostic for diabetes. If a positive diagnostic threshold for HbA1c of equal to greater than 6.1% was selected, they found that 68% of those with diabetes (sensitivity) and 98% of those without the disease (specificity) were correctly identified. Thus they take the position that this threshold is not a satisfactory stand-alone criterion for a definitive diagnosis of diabetes but rather identifies those requiring confirmatory testing. However, using a cut-off point of equal to less than 5.4%, gave the false negative rate of < 1% and identified patients who could be reliably excluded from the diagnosis of diabetes and additional tests. The range of 5.5% to 6.0% represented a grey area where it was found that HbA1c may exclude diabetics in moderate but not high-risk groups. Risk was determined by age, gender, race, hypertension, waist circumference, triglycerides and HDL cholesterol, i.e. components of the metabolic syndrome. Ginde et al point out that their cut-off of 6.1% was consistent with results from other studies.

Buell et al.[13] also used the FPG equal to greater than 126 mg/dL as the standard for diagnosis of diabetes. They suggest that if used alone, an HbA1c of equal to less than 6% could be considered normal, 6.1% to 6.9% pre-diabetic, and equal to greater than 7% diabetic. As an alternative, a value of equal to or greater than 5.8% should trigger measuring FPG or conducting an OGTT. They note, as have others that diabetic complications did not develop or progress in most Type 1 or Type 2 diabetics when the HbA1c was < 7%.

Recently Saudek et al, a group of six independent experts from the U.S., provided an up-to-date examination of the issues associated with using HbA1C for screening and diagnosis of diabetes. They point out that HbA1c does not require fasting, better reflects longer term glycemia than does plasma glucose, can be assayed with well standardized and reliable laboratory methods, and that non-glycemic factors that might interfere are rare. Based on an examination of the literature, they suggest that a HbA1C level > 6.0% should prompt further testing and follow-up, that a level of 6.5% to 6.9% or greater, confirmed by a fasting glucose or OGTT test, should establish the diagnosis of diabetes, and that a level > 7% if confirmed by just a second HbA1c test should also establish the diagnosis[14]. It is interesting in connection with the 7% cut-off that data from Peters et al 15 indicated that HbA1c equal to greater than 7% identified 99.6 % of patients with diabetes according to the OGTT standard.

It can be argued that the cut-off of 6% for the boundary between normal and abnormal or for having cause for concern may be somewhat high since apparently only a cut-off closer to 5.4% indicates a very low probability of having diabetes.

Saudek et al [14] also examine the ability of HbA1c to predict the future development of diabetes. Studies that address this issue all suffer from the fact that there are two different "gold standards" for classifying individuals as having diabetes and the agreement is modest. Nevertheless, Saudek et al cite studies that found (a) baseline HbA1c and body mass index were the only significant predictors of new onset diabetes with HbA1c the stronger; (b) FPG-defined future diabetes risk increased exponentially with baseline HbA1c; and (c) in a Japanese cohort, a baseline HbA1c equal to greater than 5.8, regardless of the FPG, carried a 10-fold increase in the risk of diagnosed diabetes over 7 years. This last result also underscores the likelihood that a cut-off for concern, additional testing and intervention in the region of 5.5% or at least somewhat less than 6% might be a good choice.

What appears to emerge from the above studies is that the three tests in question are all far from perfect and that the ability of diagnosing diabetes is improved by using a second but different test to avoid missing cases or to provide confirmation. The old approach used the OGTT a single gold standard and it was indicated if FPG suggested the risk of diabetes. Since HbA1c is a much more convenient option than the OGTT, the combination of FPG and HbA1c appears to be attractive, especially with HbA1c as the initial screening test using the protocol suggested by Saudek et al. But for HbA1c levels above 7%, Saudek et al return to the single test, if confirmed, as an adequate diagnostic for diabetes. Accepting the proposal of Saudek et al, however, goes against the current guidelines, although requests from a patient to include HbA1c in routine blood work should not be met with much resistance, given that the test is very commonly used to monitor glycemic control in diabetics[15].

It is worth reiterating that a weakness in all of these studies is having two diagnostic "gold standards" for diabetes and they not only identify somewhat different groups within a population but also even differ in which aspect of diabetic pathology they monitor. It is also clear that the current guidelines, when they yield a diagnosis of diabetes based on only one test, even if confirmed by the same test, are merely indicating that a certain definition of the disease has been satisfied, with a not insignificant probability that a different test based on a different definition might be negative. But then, given the continuous progression from normal to abnormal glucose metabolism reflected in either impaired fasting glucose or impaired glucose tolerance or both (see below), picking a point in this progression where a label of diabetes is attached is somewhat arbitrary anyway. However, the single test approach may be harmful in that it implies a definite yes or no diagnosis exists, and when it is "no" the presence of dysfunctional glucose metabolism that is almost as dangerous as diabetes may be ignored and the opportunity for highly beneficial interventions missed. However, while the above discussion suggests that there are inconsistencies and an arbitrariness in the diagnostic protocols, the other side of the coin is that criteria based on either FPG, OGTT or HbA1c, when diabetes is indicated, are with a high degree of certainty indicating an enhanced risk of microvascular associated disease which must be addressed. What is worrisome is that even in the absence of a diagnosis of diabetes, insulin resistance should be a key concern, and the only definitive way to establish its presence is with the "inconvenient" OGTT. Those reluctant to order such a test should realize that the pathological results of insulin resistance are also "inconvenient." This leads us to a discussion of pre-diabetes; the state one passes through on the way to diabetes.

According to the Centers For Disease Control, at least 25% of U.S. adults are known to have pre-diabetes but only 4% of these are actually currently diagnosed[16]. These numbers do not bode well for the future health of this population. Impaired fasting glucose and impaired glucose tolerance are the hallmarks of the pre-diabetic state. Either is sufficient to indicate pre-diabetes. The present guidelines define these two conditions as follows[2,3]

  • Impaired Fasting Glucose (IFG). WHO: FPG 110 to 125 mg/dL (6.1-6.9). ADA: FPG 100 to 125 mg/dL (5.6-6.9). WHO adds 2-hr PG < 140 mg/dL (7.8) if the OGTT is done.
  • Impaired glucose Tolerance (IGT). WHO and ADA: OGTT 2-hr plasma glucose equal to greater than 140 and < 200 mg/dL (7.8-11.1). WHO adds FPG < 126 mg/dL (7.0) as an additional condition.

The only significant difference between the WHO and the ADA is the lower limit for impaired fasting glucose. The ADA now defines normal fasting plasma glucose as < 100 mg/dL. However, WHO adds an upper limit to the 2-hour OGTT result to exclude IGT and thus identify isolated IFG. Identifying IGT requires an OGTT since IGT is related to the body's response to increased blood glucose, the normal source of which is food-derived carbohydrate. The WHO adds a FPG condition of < 126 mg/dL to their IGT criteria since exceeding this cut-off "diagnoses" diabetes which would imply IGT.

The 2008 ADA guidelines2 treat IFG and IGT as equals in the context of pre-diabetes and do not take a position as to when an OGTT is desirable. But this position ignores the fact that the physiological basis of IFG and IGT are different. The normal control of fasting glucose depends on the ability to maintain adequate basal insulin secretion and appropriate levels of insulin sensitivity in the liver to control glucose output. The OGTT measures the response to absorption of a glucose load which involves both suppressing liver glucose output and enhancing glucose uptake in the muscles and liver. This requires a prompt increase in insulin secretion and adequate liver and muscle sensitivity to insulin. IGT is associated mostly with peripheral muscle insulin resistance, i.e. lack of insulin sensitivity.

If either 110 or 100 mg/dL for FPG is regarded as normal and >126 mg/dL is considered diagnostic for diabetes, then there is an intermediate state in between which is called IFG and which is viewed as a pre-diabetic state. However, as discussed above, this of necessity presents only a partial picture of an individual's glucose metabolism since it is based exclusively on the fasting state. An equally important if not more important problem in the context of pre-diabetes is how the body handles ingested glucose, mostly in the form of carbohydrate, and whether the resultant swings in post meal blood glucose are normal or abnormal. Independent of whether or not an individual falls in the IFG category, it would seem highly desirable to investigate the possibility of IGT, and the standard approach is and always has been the OGTT, as illustrated in the above guidelines. To determine how well a patient handles food-derived blood glucose, the only way outside of a research setting to obtain an answer is to provide the patient in a fasting state with a glucose challenge and see what happens. The point is that some who have IFG will also have IGT, others will not, and some who have IGT will have IFG, others will not. Thus there is also the problem of identifying those with isolated IGT and if a normal FPG terminates any further investigation of glucose metabolism, isolated IGT will be missed. Given that IGT is a marker for the pre-diabetic state and that in the context of diabetes prevention, knowledge of its presence is important, it seems that avoiding the "inconvenient" OGTT if the FPG is normal is counter productive.

What is really at issue here is the question of the presence of insulin resistance and this is much more difficult to establish than IFG. Gerald Reaven, who some would call the father of the metabolic syndrome, has recently pointed out that "it is not clear that the use of IFG provides a particularly effective way to identify either the presence of insulin resistance or to predict CVD risk."[17] Reaven and coworkers studied a group of 490 men and women free of any disease and not taking any drug that influenced carbohydrate metabolism. All were subjected to an insulin resistance evaluation based on measuring insulin-mediated glucose disposal, a complex procedure suited only to the research setting. This allowed the classification of the group into tertiles, the lowest third representing insulin sensitive and the highest those classified as being strongly insulin resistant. Then the ADA criteria given above were applied to classify the individuals with regard to having normal glucose tolerance (NGT). Out of 404 classified as having NGT, 104 were in the insulin resistant tertile. Thus 26% of those who were the most insulin resistant were missed by the fasting glucose criterion and would not have been considered pre-diabetic even though their insulin resistance was high and carried an unfavourable prognosis.[18]

Another example is provided by a recent study.[5] Two hundred and twenty patients with a FPG between 110 and 125 mg/dL, i.e. IFG were evaluated by an OGTT. It was found that 32.8% had IGT, 33.6% also had Diabetes and only 33.6% had normal glucose tolerance. The missed diabetics in this study have been discussed above. The measurement of fasting glucose provided an incomplete picture of the glucose metabolism in this cohort.

Thus it is interesting to examine what role HbA1c might play in this context. This was investigated by Geverhiwot et al [19]. A group of 225 patients with FPG equal to less than 108 mg/dL (6.0), which by earlier guidelines was normal, underwent an OGTT and an HbA1c measurement. Of these, 45 had IGT (20%) and 7 had diabetes (3%) by the WHO criteria. Subjects with abnormal glucose tolerance had higher HbA1C levels and a cut-off of 5.6% gave optimal sensitivity and specificity to predict a 2-hour glucose of > 140 mg/dL (7.8), i.e. IGT. Note that a low FPG did not rule out IGT. While this was a small study, the result that 20% of those with a FPG < 108 mg/dL had IGT suggests that if one really wants to know about the status of their carbohydrate and glucose metabolism, then even when the FPG is at the high end of normal (or just above it by the new ADA guidelines), obtaining a HbA1C value can provide additional valuable information and could justify an OGTT to get a definitive answer. Also, it calls into question the use of an HbA1c cut-off of 6.0 or 6.1% below which on need not have any concerns. Being concerned about pre-diabetes is just as important as being concerned about diabetes.

The diagnosis of the condition called the metabolic syndrome should also be a wake-up call even if the fasting glucose is below the cut-off (< 100 mg/dL used as a factor in the diagnosis of this disorder). Other factors in the metabolic syndrome definition are strongly correlated with insulin resistance, which is in fact considered to be at the root of the adverse outcomes associated with this syndrome.[20]

Epidemiologic studies have consistently demonstrated a positive association between HbA1c levels and cardiovascular disease in patients both with and without diabetes.[21-26] In fact, HbA1c levels appear to be an independent risk factor for CHD and CVD and also account for the failure of conventional risk factors to explain all of the enhanced risk associated with diabetes.[22] There are several proposed mechanisms which link hyperglycemia with adverse cardiovascular outcomes. These include inflammatory cytokines, endothelial dysfunction and hypercoagulability in the setting of platelet activation which initiates a cascade of events accelerated by hyperglycemia.[22]

In one of the largest prospective follow-up studies (the Epic-Norfolk study) 4662 men and 5570 women between the ages of 45 and 79 were assessed from 1995 to 1997 and followed until 2003. Most were non-diabetics. Persons with HbA1c < 5% had the lowest rates of CVD and mortality. An increase of 1 percentage point was associated with an increase in risk of death from any cause of 24% in men and 28% in women. These risk elevations were independent of age, body mass index, waist-to-hip ratio, systolic blood pressure, cholesterol levels, smoking and a history of CVD.25 The Epic-Norfolk study also examined HbA1c levels and mortality in diabetics. It was found that glycated haemoglobin levels explained most of the excess mortality risk associated with diabetes in men and was a continuous, positive risk factor.[27]

The strong and consistent association between CVD and HbA1c levels indicates that concern over levels above around 5% and especially equal to greater than 6% should be motivated by the recognition that hyperglycemia is not only a risk factor for diabetes and its multiple microvascular complications, but also for CVD in general. This would seem to provide a strong argument for its routine measurement during physical examinations, even in the absence of risk factors for diabetes.

Hyperglycemia also appears to impact the risk and progression of several common cancers. In fact, the notion that "cancer loves sugar" is part of long-standing folklore. Positive associations have been found between FPG and liver cancer risk, and the incidence of non-Hodgkin's lymphoma, colorectal cancer, bladder cancer, thyroid cancer, multiple myeloma and breast cancer after age 65.[28] Elevated HbA1C (>7.5%) has been identified as an independent predictor of clinically aggressive colorectal cancer in patients with Type 2 diabetes.[29]

Knowing that one has impaired fasting glucose, impaired glucose tolerance or an elevated HbA1c is of little use unless there are interventions that will normalize hyperglycemia. This is a problem that faces not only pre-diabetics but also diabetics. Type 1 diabetics attempt to control glucose levels with diet and insulin injections. Many Type 2 diabetics do not require insulin and are advised to exercise, lose weight, perhaps change certain dietary practices and even take prescription medicine designed to aid in glucose control. Pre-diabetics may be offered similar dietary and exercise advice or the problem may be simply be ignored by a medical culture that concentrates on "real diseases" which it treats mostly with pharmaceuticals. In fact, in a 2007 statement on IFG and IGT which carried the subtitle "Implications for care," the ADA fails to mention dietary details or modifications at all and simply recommends exercise, weight loss or drugs (e.g. metformin), thereby ignoring considerable research regarding the dietary aspect of hyperglycemia.[30]

Dietary and lifestyle changes that may prove beneficial when glucose metabolism is defective have been discussed in another research report Coronary Heart Disease Risk and Its Reduction. While that review was concerned with the reduction of the risk of coronary heart disease, the same principles and actions are applicable to the problems discussed in this review. The use of diet and exercise for the prevention of coronary heart disease applies equally well to the prevention of diabetes and insulin resistance and in fact, the review on coronary heart disease risk and its reduction deals at length reducing the risk of diabetes and the metabolic syndrome. The risk reductions of 80-90% in the incidence of diabetes obtained in the prospective studies cited, which were through diet and exercise, provide confirmation that the selection of carbohydrates and the dietary glycemic load are indeed critical factors. The success of dietary and lifestyle interventions in the context of IGF and IGT can be judged by periodic HbA1c determinations. In fact, this may be the best way to assess intervention progress. It is important to recognize that IFG and IGT are intermediate states along the path to diabetes and preventing diabetes a priori means arresting the progression of IGT and IFG ideally reversing these metabolic abnormalities.

There appear to be only three randomized trials of changes in lifestyle and diet that addressed individuals with IGT.[31-33] In terms of trials, IFG appears ignored. These studies provided evidence that such changes can dramatically reduce the incidence of diabetes and reduce HbA1c progression or reverse its progression in non-diabetic individuals. Dietary interventions in these studies, which reported in the period 1997 to 2002, were influenced by the "fat is dangerous to your health" dogma and thus it is possible that even more dramatic results could have been obtained by using judicious choice of carbohydrate type and quantity to limit post-meal blood glucose swings. While low-fat diets may have helped some to reduce weight, as discussed in the above mentioned research review, this strategy frequently enhanced other symptoms of the metabolic syndrome.

The book Dr. Bernstein's Diabetes Solution, The complete Guide to Achieving Normal Blood Sugars (Little Brown, New York, 2003) should be of interest to anyone with blood glucose problems, even if not diabetic. Bernstein is somewhat of a maverick M.D. and a strong believer in carbohydrate restriction and selection for blood glucose control that goes way beyond the ADA recommendations. He is a Type 1 diabetic himself who specializes in diabetes. Patients come to his clinic from all over the world. His book provides useful information on the use of diet to control post-meal hyperglycemia and normalize HbA1c levels in both diabetics and pre-diabetics. His aim is to achieve in his patients HbA1c levels similar to lean non-diabetics, which are in the range of 4.2% to 4.6%. This is in sharp contrast to the target of 7% advocated by the ADA for diabetics. He regards the 7% target as "out of control" rather than tight control, the ADA view. He considers the HbA1c levels associated with IFG and IGT to also be dangerously high if micro- and macrovascular damage is the issue, i.e. he believes the threshold for long-term vascular damage is lower than the generally believed.

Readers of International Health News will be aware of the constant stream of studies which are concerned with reducing the risk of heart disease and diabetes. The general theme that emerges in all these studies is the avoidance of refined grains and sugar and the emphasis on fruits, vegetables, beans, nuts and fish, moderation in red and processed meat consumption, weight control or reduction and exercise. The importance of preventing hyperglycemia in the context of both heart disease and diabetes is highlighted in many studies that emphasize low glycemic load diets or low glycemic index diets and as well, the Mediterranean diet. It is becoming clearer by the day that defective glucose metabolism and hyperglycemia represent a serious threat to living a long, healthy life.[34]

At issue here is the risk of serious eye and kidney disorders, peripheral circulation problems and the risk of amputation, coronary heart disease and stroke. It seems clear that hyperglycemia even in non-diabetics should not be ignored. A fasting glucose of > 100-124 mg/dL (5.6-6.9 mmol/L) or an HbA1c greater than about 5.5% should prompt concern and further testing. Impaired glucose tolerance is much more prevalent than impaired fasting glucose and can only be identified by a 2-hour oral glucose tolerance test. Diet, weight loss and exercise represent a trio of interventions that have been shown to strongly impact the progression to diabetes. This of necessity translates to the prevention of progression of impaired glucose tolerance and impaired fasting glucose to the cut-offs for the" diagnosis" of diabetes. These are the abnormal intermediate states, and the reversal of the progression of these two disorders prevents diabetes. Since post-meal blood glucose fluctuations impact the average blood glucose level, the carbohydrate content of the diet becomes a central issue although it appears to be downplayed by mainstream medicine which still seems obsessed with dietary fat.

It is important to recognize that dietary carbohydrate may well be the key to the whole problem, although weight loss and subsequent control may also be critical. Carbohydrate addiction or heavy carbohydrate consumption, which are very common, especially if one is advised to limit fat, results in fat storage which results in weight gain which results in increased insulin resistance which results in hyperglycemia. The end result is obesity and pre-diabetes leading to diabetes. Not everyone agrees. Some consider carbohydrates benign. The ADA suggests that less than 120 g/day is not advisable. Bernstein, who tries to regulate blood sugar levels in both diabetics and pre-diabetics, suggests 30 g/day and in some cases achieves HbA1c levels well below 5%, even in diabetics! Most individuals would find 30 g/day unacceptable if not impossible except over a fairly short term. Such extreme restriction is also probably not necessary to normalize FPG and the OGTT in many individuals. Strongly elevated post-meal glucose levels are almost invariably the result of heavy carbohydrate consumption, especially rapidly digested carbohydrates (from high glycemic index foods such as most bread, rice, potatoes, carrots, corn, candy, many breakfast cereals and most deserts). They in fact tend to make one feel good and the physiology is well understood. Finally, HbA1c can be a useful measure of progress in glucose control even in non-diabetics, although for individuals with IGT, it seems reasonable that success should be measured by a return to a normal OGTT, independent of FPG, given that the absence of insulin resistance is the ultimate goal.

The reader is also referred to the research report Carbohydrate Restriction for a discussion of the role of restricting carbohydrates in the diet and exercise in normalizing glucose metabolism in Type 2 diabetics. The Studies of Westman, Vernon and coworkers are particularly pertinent and apply equally well- to prediabetics.[35,36]

  1. Ealovega MW, Tabaei BP, Brandle M, Burke R, Herman WH. Opportunistic screening for diabetes in routine clinical practice. Diabetes Care 2004 January;27(1):[9-12].
  2. American Diabetes Association. Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 2008 January 1;31(Supplement_1):S55-S60.
  3. The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus*. Follow-up Report on the Diagnosis of Diabetes Mellitus. Diabetes Care 2003 November 1;26(11):3160-7.
  4. DECODE Study Group. Will new diagnostic criteria for diabetes mellitus change phenotype of patients with diabetes? Reanalysis of European epidemiological data. BMJ 1998 August 8;317(7155):371-5.
  5. Tutuncuoglu P, Sarac F, Saygili F, Ozgen AG, Yilmaz C, Tuzun M. Diabetes and impaired glucose tolerance prevalences in Turkish patients with impaired fasting glucose. Acta Diabetol 2008 September;45(3):151-6.
  6. Kernohan AF, Perry CG, Small M. Clinical impact of the new criteria for the diagnosis of diabetes mellitus. Clin Chem Lab Med 2003 September;41(9):1239-45.
  7. Troisi RJ, Cowie CC, Harris MI. Diurnal variation in fasting plasma glucose: implications for diagnosis of diabetes in patients examined in the afternoon. JAMA 2000 December 27;284(24):3157-9.
  8. Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ. Translating the A1C assay into estimated average glucose values. Diabetes Care 2008 August;31(8):1473-8.
  9. Rohlfing CL, Wiedmeyer HM, Little RR, England JD, Tennill A, Goldstein DE. Defining the Relationship Between Plasma Glucose and HbA1c: Analysis of glucose profiles and HbA1c in the Diabetes Control and Complications Trial. Diabetes Care 2002 February 1;25(2):275-8.
  10. Hassan Y, Johnson B, Nader N, Gannon MC, Nuttall FQ. The relationship between 24-hour integrated glucose concentrations and % glycohemoglobin. J Lab Clin Med 2006 January;147(1):21-6.
  11. Bennett CM, Guo M, Dharmage SC. HbA(1c) as a screening tool for detection of Type 2 diabetes: a systematic review. Diabet Med 2007 April;24(4):333-43.
  12. Ginde AA, Cagliero E, Nathan DM, Camargo CA, Jr. Value of risk stratification to increase the predictive validity of HbA1c in screening for undiagnosed diabetes in the US population. J Gen Intern Med 2008 September;23(9):1346-53.
  13. Buell C, Kermah D, Davidson MB. Utility of A1C for diabetes screening in the 1999 2004 NHANES population. Diabetes Care 2007 September;30(9):2233-5.
  14. Saudek CD, Herman WH, Sacks DB, Bergenstal RM, Edelman D, Davidson MB. A new look at screening and diagnosing diabetes mellitus. J Clin Endocrinol Metab 2008 July;93(7):2447-53.
  15. Peters AL, Davidson MB, Schriger DL, Hasselblad V. A clinical approach for the diagnosis of diabetes mellitus: an analysis using glycosylated hemoglobin levels. Meta-analysis Research Group on the Diagnosis of Diabetes Using Glycated Hemoglobin Levels. JAMA 1996 October 16;276(15):1246-52.
  16. Self-reported prediabetes and risk-reduction activities--United States, 2006. MMWR Morb Mortal Wkly Rep 2008 November 7;57(44):1203-5.
  17. Reaven GM. The metabolic syndrome: requiescat in pace. Clin Chem 2005 June;51(6):931-8.
  18. Tuan CY, Abbasi F, Lamendola C, McLaughlin T, Reaven G. Usefulness of plasma glucose and insulin concentrations in identifying patients with insulin resistance. The American Journal of Cardiology 2003 September 1;92(5):606-10.
  19. Geberhiwot T, Haddon A, Labib M. HbA1c predicts the likelihood of having impaired glucose tolerance in high-risk patients with normal fasting plasma glucose. Ann Clin Biochem 2005 May;42(Pt 3):193-5.
  20. Ware WR. Psychological stress, insulin resistance, inflammation and the assessment of heart disease risk. Time for a paradigm shift? Med Hypotheses 2008;71(1):45-52.
  21. ten BR, Dekker N, de GM, Ikkersheim D. Lowering HbA1c in type 2 diabetics results in reduced risk of coronary heart disease and all-cause mortality. Prim Care Diabetes 2008 February;2(1):45-9.
  22. Golden SH, Selvin E, Cunningham KE. Glycaemic status and cardiovascular disease in type 2 diabetes mellitus: re-visiting glycated haemoglobin targets for cardiovascular disease prevention. Diabetes Obes Metab 2007 November;9(6):792-8.
  23. Selvin E, Coresh J, Golden SH, Brancati FL, Folsom AR, Steffes MW. Glycemic control and coronary heart disease risk in persons with and without diabetes: the atherosclerosis risk in communities study. Arch Intern Med 2005 September 12;165(16):1910-6.
  24. Gerstein HC, Swedberg K, Carlsson J et al. The hemoglobin A1c level as a progressive risk factor for cardiovascular death, hospitalization for heart failure, or death in patients with chronic heart failure: an analysis of the Candesartan in Heart failure: Assessment of Reduction in Mortality and Morbidity (CHARM) program. Arch Intern Med 2008 August 11;168(15):1699-704.
  25. Khaw KT, Wareham N, Bingham S, Luben R, Welch A, Day N. Association of hemoglobin A1c with cardiovascular disease and mortality in adults: the European prospective investigation into cancer in Norfolk. Ann Intern Med 2004 September 21;141(6):413-20.
  26. Selvin E, Marinopoulos S, Berkenblit G et al. Meta-analysis: glycosylated hemoglobin and cardiovascular disease in diabetes mellitus. Ann Intern Med 2004 September 21;141(6):421-31.
  27. Khaw KT, Wareham N, Luben R et al. Glycated haemoglobin, diabetes, and mortality in men in Norfolk cohort of European prospective investigation of cancer and nutrition (EPIC-Norfolk). BMJ 2001 January 6;322(7277):15-8.
  28. Rapp K, Schroeder J, Klenk J et al. Fasting blood glucose and cancer risk in a cohort of more than 140,000 adults in Austria. Diabetologia 2006 May;49(5):945-52.
  29. Siddiqui AA, Spechler SJ, Huerta S, Dredar S, Little BB, Cryer B. Elevated HbA1c is an independent predictor of aggressive clinical behavior in patients with colorectal cancer: a case-control study. Dig Dis Sci 2008 September;53(9):2486-94.
  30. Nathan DM, Davidson MB, DeFronzo RA et al. Impaired fasting glucose and impaired glucose tolerance: implications for care. Diabetes Care 2007 March;30(3):753-9.
  31. Pan XR, Li GW, Hu YH et al. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care 1997 April 1;20(4):537-44.
  32. Tuomilehto J, Lindstrom J, Eriksson JG et al. Prevention of Type 2 Diabetes Mellitus by Changes in Lifestyle among Subjects with Impaired Glucose Tolerance. N Engl J Med 2001 May 3;344(18):1343-50.
  33. Knowler WC, Barrett-Connor E, Fowler SE et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002 February 7;346(6):393-403.
  34. Giugliano D, Ceriello A, Esposito K. Glucose metabolism and hyperglycemia. Am J Clin Nutr 2008 January;87(1):217S-22S.
  35. Westman EC, Vernon MC. Has carbohydrate-restriction been forgotten as a treatment for diabetes mellitus? A perspective on the ACCORD study design. Nutr Metab (Lond) 2008;5:10.
  36. Vernon MC, Mavropoulos J, Transue M, Yancy WS, Westman EC. Clinical experience of a carbohydrate-restricted diet: effect on diabetes mellitus. Metab Syndr Relat Disord 2003 September;1(3):233-7.
This article was first published in the December 2008/January 2009 issue of International Health News


Subscription to IHN

Donation appeal

My favourite Supplements

copyright notice