Beyond HbA1c and glucose: the role of nontraditional glycemic markers in diabetes diagnosis, prognosis, and management

Abstract

Fasting glucose and hemoglobin A1c (HbA1c) are the standard measures for diagnosis and monitoring of diabetes. There has been recent interest in nontraditional markers of hyperglycemia, including fructosamine, glycated albumin and 1,5-anhydroglucitol (1,5-AG), as alternatives or adjuncts to standard measures. There is a growing literature linking these nontraditional markers with microvascular and macrovascular complications. Fructosamine and glycated albumin have also been shown to improve identification of persons with diabetes. However, long-term prospective studies with clinical outcomes are lacking. Some modern laboratory assays for fructosamine, glycated albumin and 1,5-AG have excellent performance. Expanded use of these tests has the potential to improve diabetes care as these measures may overcome limitations of HbA1c in certain patients, complement traditional measures by providing additional information on shorter-term glycemic control, and improve risk stratification for diabetes and its complications. Nonetheless, studies are needed to demonstrate if their routine use will benefit patients and improve outcomes.

Introduction

Hemoglobin A1c (HbA1c) has long been the standard measure used to monitor glycemic control in clinical practice and is routinely measured in all persons with diabetes. In addition to fasting glucose and 2-hour glucose, the American Diabetes Association (ADA), European Association for the Study of Diabetes (EASD), the World Health Organization (WHO), and other diabetes organizations now recommend the use of HbA1c for diagnosis of diabetes.^1–6 First recommended in 2009, the addition of HbA1c to diagnostic criteria for diabetes has been controversial, largely attributable to limitations of the HbA1c test.^7–9 There is growing interest in serum biomarkers of hyperglycemia, including fructosamine, glycated albumin, and 1,5-anhydroglucitol (1,5-AG), to be used as alternatives to or in conjunction with traditional measures^10–13 These markers can overcome limitations of HbA1c in certain patients, could complement traditional measures in the clinic by providing additional information on shorter-term glycemic control, and may improve risk stratification for diabetes and its complications.

Markers of hyperglycemia

Traditional markers of hyperglycemia

Diabetes is a condition defined by elevations in glucose. Historically, glucose measured in the fasting state or glucose measured two hours after a carbohydrate challenge (oral glucose tolerance test) have been the standard measures used to diagnose diabetes and identify people at risk for diabetes (frequently termed “prediabetes”). HbA1c has been used widely since the 1980s and is the standard measure used for monitoring glycemic control in clinical practice.⁵ In red blood cells, HbA1c is hemoglobin that has glucose attached to the N-terminal valine of the beta chain and is reported as the proportion of total hemoglobin. Because the lifespan of red blood cells is approximately 120 days, HbA1c therefore reflects average glycemia over the past two to three months (since it is weighted towards the more recent months).¹³ Advantages of HbA1c include the lack of participant preparation (fasting is not necessary); high within-person reliability;^14,15 and excellent standardization of the assay in most countries.^16–18 Nonetheless, disadvantages of HbA1c include limited interpretability in the setting of altered red blood cell lifespan—levels are affected by changes in duration of red blood cell exposure to circulating blood glucose levels—and interference of some HbA1c assays by hemoglobin variants and several rare conditions (Table 1).^7,19 These disadvantages have brought into focus possible roles for nontraditional glycemic markers in the clinic.

Table 1.

Characteristics of traditional and nontraditional markers of hyperglycemia

	Brief Description	Duration of glycemia reflected	Strengths	Limitations
Traditional markers of hyperglycemia
Fasting glucose	Direct measure of circulating blood glucose	Acute/immediate	Direct measure Widely accepted Inexpensive	Requires fasting; affected by acute illness and stress; pre-analytical issues (sample stability)*; moderate within-person variability
HbA1c	Proportion of hemoglobin that is glycated	2–3 months	Reflects 2–3 month control Low within-person variability; no patient preparation needed; not affected by acute illness, stress or recent activity levels	Affected by alterations in red cell turnover; some methods for measurement can give inaccurate results in the presence of certain hemoglobin variants**; requires whole blood; cost
Nontraditional markers of hyperglycemia
Fructosamine	Total serum protein glycation	2–3 weeks	Does not require fasting; highly reliable automated methods are widely available; can be measured in serum or plasma; inexpensive	Affected by changes in serum protein metabolism (mostly albumin), thyroid dysfunction; limited evidence linking to outcomes
Glycated albumin	Proportion of albumin that is glycated	2–3 weeks	Does not require fasting; can be measured in serum or plasma	Affected by changes in albumin metabolism, thyroid dysfunction; method performance may vary; availability in the US is limited; limited evidence linking to outcomes
1,5-AG	Monosaccharide filtered by the kidney and normally reabsorbed; reabsorption inhibited and it is excreted at high levels of glycemia, so serum levels drop	2–14 days	Does not require fasting; can be measured in serum or plasma; test is available from major laboratories in the US; expense	Affected by changes in renal threshold for glucose, dialysis or stage 4 or 5 kidney disease, pregnancy; limited evidence linking to outcomes

^*See: Gambino R. Clin Chem. 2007 Dec;53(12):2040-1.

^**See: www.ngsp.org for comprehensive list

Nontraditional markers of hyperglycemia

Fructosamine and glycated albumin are both ketoamines, which are formed as the result of a non-enzymatic process that binds glucose to serum proteins. In states of abnormally high glucose concentrations, as in persons with diabetes, serum proteins are exposed to greater concentrations of glucose and therefore experience increased glycation.²⁰ Fructosamine assays measure total glycated serumprotein (mostly albumin, but also immunoglobulins and other circulating proteins), whereas glycated albumin is reported as the proportion of total albumin. The half-life of albumin and other serum proteins is shorter than that of red blood cells; thus measurements of fructosamine and glycated albumin reflect average glycemia over a shorter duration, approximately two to three weeks.²⁰

1,5-AG is a 6-carbon monosaccharide obtained mainly from dietary sources, that reflects average glycemia over approximately the past 2–14 days.^21–25 In states of normal glycemia, nearly 100% of 1,5-AG is reabsorbed by the renal tubule. However, at very high levels of glycemia (above the renal threshold, ~160–180 mg/dl), glucose competes with 1,5-AG for reabsorption by the renal tubule, and 1,5-AG is excreted in the urine, resulting in a drop in circulating 1,5-AG levels in the blood. Therefore, there is an inverse association between high levels of glucose and 1,5-AG²¹. Soybeans have particularly high levels of 1,5-AG, and certain foods such as rice, bread and beef contain modest levels; it is unclear to what extent dietary intake may affect circulating 1,5-AG levels and the interpretation of this test.^21,22

Correlations of traditional markers of hyperglycemia with fructosamine, glycated albumin, and 1,5-anhydroglucitol

Fructosamine and glycated albumin are strongly associated with HbA1c and fasting glucose,^26–30 and all four measures have been shown to be similarly correlated with mean glucose from continuous glucose monitoring over about 5 days in persons with diabetes.³¹ In settings where HbA1c testing is known to be problematic, fructosamine or glycated albumin may be a useful substitute. A difficulty, however, is that there are no established clinical cut-points and these assays are not standardized across instruments. Conversion equations can help estimate the ranges of fructosamine and glycated albumin test results that are similar to HbA1c targets. Various equations have been developed to convert fructosamine and glycated albumin to an “HbA1c equivalent”. For example, previous reports demonstrated that glycated albumin values in the range of 16% to 22%,^27,32–34 and fructosamine levels around 312 μmol/L as reported by one study,²⁷ are approximately equivalent to an HbA1c value of 7%. 1,5-AG is strongly inversely associated with HbA1c and fasting glucose in persons with diagnosed diabetes,²⁷ but appropriate clinical targets are unclear. It should be noted that 1,5-AG is poorly correlated with fasting glucose and HbA1c in persons without diagnosed diabetes--the strongest correlations are observed at the highest glucose concentrations.²⁷ (additionally cite Selvin in press) This suggests the utility of 1,5-AG may primarily be limited to persons with overtly elevated glucose.

Since these markers of hyperglycemia are measured on different scales, both clinicians and patients may benefit from being provided with equivalents. However, conversion equations for nontraditional glycemic markers have typically relied on single measurements (which may vary considerably over time, particularly in diabetic patients) and may differ depending on the underlying population from which they are derived, with uncertain generalizability. Furthermore, none of these markers are perfectly correlated, a function of differences in the physiology of each biomarker including the duration of glycemia reflected and other sources of biological and analytical variability. In fact, the discordance across traditional and nontraditional glycemic markers may suggest the complementary nature of these biomarkers. A benefit to the use of multiple measures is that they may each provide unique insight into different aspects of hyperglycemia and diabetes physiology.

Associations of nontraditional markers of hyperglycemia with complications

Cross-sectional studies

Cross-sectional studies have linked nontraditional markers of hyperglycemia with both microvascular and macrovascular complications. Fructosamine and glycated albumin have both been linked to prevalent retinopathy.^35–37 In a recent analysis of 12,306 persons (958 with diabetes) in the Atherosclerosis Risk in Communities (ARIC) Study, we found an independent association of glycated albumin and fructosamine with retinopathy, with patterns of association very similar to those observed for HbA1c (Figure).²⁹ In a Japanese cohort of more than 2,500 participants, the performance of glycated albumin and 1,5-AG to identify cases of retinopathy—as measured by the C-statistic—was shown to be comparable to fasting glucose and HbA1c.³⁸ In a study of 1,575 Japanese adults without diagnosed diabetes, glycated albumin was associated with carotid artery intima-media thickness, a measure of subclinical atherosclerosis.³⁹ Glycated albumin has been associated with prevalent kidney outcomes,^40–42 and cardiovascular disease.^42–49 Few studies have assessed the relationship of 1,5-AG to complications, although lower 1,5-AG concentrations have been linked to both prevalent coronary heart disease ⁵⁰ and retinopathy⁵¹ in persons with diabetes. 1,5-AG has also been associated with measures of atherosclerosis and cardiovascular disease in a population without a history of diabetes.⁵²

Figure.

Prevalence of retinopathy by deciles of fructosamine, glycated albumin, HbA1c, an fasting glucose, the Atherosclerosis Risk in Communities Study, N=9,445

Source: Selvin E, Rawlings AM, Grams M, et al. Fructosamine and glycated albumin for risk stratification and prediction of incident diabetes and microvascular complications: a prospective cohort analysis of the Atherosclerosis Risk in Communities (ARIC) study. Lancet Diabetes Endocrinol. 2014;2(4):279–288. doi:10.1016/S2213-8587(13)70199-2.

Prospective studies

Limited evidence from prospective studies suggests nontraditional markers may be useful for identification of persons at risk of developing microvascular and macrovascular complications. In addition to the associations with retinopathy in the above-mentioned ARIC Study, we found that both fructosamine and glycated albumin strongly predicted incident chronic kidney disease (CKD) over two decades of follow-up. The observed associations of fructosamine and glycated albumin with incident CKD were of similar magnitude to those observed for HbA1c.²⁹ Analyses conducted in the DCCT/EDIC study of persons with type 1 diabetes also reported that glycated albumin was similarly associated with retinopathy and nephropathy as compared to HbA1c.²⁸ Additionally, in an analysis of 84 persons with type 1 diabetes from the Wisconsin Diabetes Registry Study, fructosamine was associated with incident retinopathy.⁵³ By contrast, in a Brazilian cohort of persons with diabetes, fasting glucose was associated with microvascular outcomes over about 5 years of follow-up, but fructosamine was not.⁵⁴ In a recent prospective study in 2,095 Japanese persons (including approximately 100 with diabetes), 1,5-AG was associated with incident cardiovascular events during 11 years of follow-up.⁵⁵

Clinical utility of nontraditional markers of hyperglycemia

For monitoring of short-term glycemic control

Nontraditional markers of hyperglycemia are not formally incorporated into clinical guidelines in the United States. However, various organizations in multiple countries, including the US, India, Australia and the United Kingdom, have suggested fructosamine as a useful alternative to HbA1c for monitoring glycemic control in persons with conditions that may interfere with the interpretation of the HbA1c test.^{11,12,56–61} Glycated albumin is used frequently in China, Japan and South Korea for monitoring intermediate glycemic control.⁶² Several assays have been developed to measure glycated albumin but the assays are not standardized, and therefore not necessarily equivalent. Some early studies raised serious concerns regarding the validity and reliability of fructosamine assays⁶³, although second-generation assays had improved technical performance.⁶⁴ Modern automated assays for fructosamine have shown high correlations with glucose and HbA1c, strong prognostic value, and very low CVs (approximately 3% in recent studies)^29,31,65.

Whereas HbA1c reflects long-term, 2–3 month glycemic control, fructosamine and glycated albumin reflect hyperglycemia over the past 2 to 3 weeks. Thus, both have been proposed as useful markers of intermediate glycemic control. In clinical practice, HbA1c is typically measured at minimum every 6 months and more frequently (quarterly) in persons with recent therapy changes who are not meeting treatment goals.^1,66

Fructosamine and glycated albumin may be quite useful to evaluate earlier response to changes in treatment. Glycated albumin has been shown to change faster than HbA1c in response to changes in medication or exercise.^67,68 Compared to HbA1c, glycated albumin is more strongly correlated with continuous glucose measurements over 1 to 2 days, ^69,70 and may more accurately reflect long-term glycemic variability and glucose excursions.^71,72

1,5-AG is thought to reflect hyperglycemia over the past 2 weeks and is recommended by the manufacturer for use in persons with diabetes and HbA1c <8% to help identify patients with frequent hyperglycemic excursions.^73,74 Indeed, 1,5-AG has been shown to be correlated with postprandial hyperglycemia in persons with diabetes and HbA1c <7%;⁷⁵ and to be more strongly correlated with glucose variability as compared to HbA1c, fructosamine or glycated albumin over 2 to 3 days in persons with moderate glycemic control (HbA1c <8%).^76,77

For diabetes screening or diagnosis

There is evidence that nontraditional markers of hyperglycemia may help to more accurately identify persons with diabetes. In several studies, fructosamine and glycated albumin had similar performance for the identification of persons with diabetes as compared to either fasting glucose or HbA1c.^{27,30,78–80} Furthermore, compared to using either test individually, sensitivity to identify cases of diabetes defined by 2-hour glucose was improved when glycated albumin was used in combination with either fasting glucose or HbA1c.^80,81

A large proportion of persons identified as having pre-diabetes do not go onto develop diabetes, highlighting the need for strategies that will accurately identify persons who will progress to overt diabetes.⁸² It is possible that fructosamine or glycated albumin may be useful in early identification of high-risk persons. Recent studies have shown that both fructosamine and glycated albumin are associated with future risk of diabetes, independent of fasting glucose and HbA1c.^29,83 1,5-AG has also been associated with future development of diabetes, but observed associations were lower in magnitude as compared to other markers of hyperglycemia and were not present in persons with fasting glucose or HbA1c in the non-diabetic range.⁸³ Nonetheless, the evidence linking nontraditional biomarkers with future diabetes risk is sparse.

Utility of nontraditional markers in special populations

A focus in the literature has been the potential utility of fructosamine or glycated albumin for monitoring glycemic control in the setting of certain populations where HbA1c is thought to inaccurately reflect glycemia, including severe kidney disease.⁸⁴ Recent studies have shown that, compared to HbA1c, glycated albumin is more strongly correlated with glucose in dialysis patients.^85–92 Fructosamine and glycated albumin may also be useful for prediction of complications in persons with kidney failure. Indeed, fructosamine and glycated albumin have been both cross-sectionally and prospectively associated with microvascular, macrovascular and all-cause morbidity and mortality in dialysis patients, whereas many studies have reported no association of HbA1c with these outcomes.^65,93–100 Nonetheless, despite their associations with clinical outcomes, fructosamine and glycated albumin may also be limited in this setting, since proteinuria and altered serum protein turnover may affect interpretation of these tests.^101–103

1,5-AG has not been well studied in the setting of chronic kidney disease or dialysis. Because lowered plasma concentrations of 1,5-AG result from accelerated urine excretion due to competitive inhibition of glucose by the renal tubules, 1,5-AG may have a problematic interpretation in the setting of reduced kidney function. 1,5-AG was correlated with fasting glucose and HbA1c in persons with diabetes and mild to moderate CKD, but not in those with end stage renal disease (ESRD) (stages 4–5 CKD).¹⁰⁴

There is also evidence to support the use of nontraditional markers of hyperglycemia in persons with other conditions that may decrease the lifespan of red blood cells. Fructosamine and glycated albumin have been shown to better reflect glucose levels in the setting of anemia, autologous blood donations and HIV, which may all result in artificially low HbA1c.^105–108 There is also interest in whether fructosamine, glycated albumin, or 1,5-AG testing may play a role in the management of diabetes in patients with liver disease, but evidence for their performance in this setting is inconsistent.^109–111 Furthermore, during pregnancy, glycated albumin may better reflect average glucose compared to HbA1c, which may be artificially elevated due to iron deficiency.^112,113 Furthermore, measures of shorter-term glycemia may be especially important in gestational diabetes given the importance of frequent monitoring and strong associations between diabetes control in pregnancy and maternal and fetal outcomes.^1,114

Conclusions

Nontraditional markers of hyperglycemia, fructosamine, glycated albumin, or 1,5-AG, may be useful for monitoring of glycemic control when short-term changes are of interest or as alternatives to HbA1c in settings in which HbA1c may be problematic. Fructosamine and glycated albumin may also aid in early identification of persons at future risk for diabetes. In clinical or epidemiologic studies where fasting glucose or HbA1c measurements are not available but where serum or plasma specimens were collected, fructosamine or glycated albumin may be particularly useful to identify persons with undiagnosed hyperglycemia. Furthermore, in resource-intensive randomized clinical trials of short duration (<6 months), fructosamine and glycated albumin may be useful to evaluate responses to glucose-lowering interventions. Additionally, the complementary nature of these different tests of hyperglycemia warrants exploration into the potential utility of fructosamine, glycated albumin, and/or 1,5-AG in the development of risk prediction models for diabetes and its complications.

Additional studies of fructosamine, glycated albumin and 1,5-AG could help address uncertainty in this area. First, prospective associations of these three nontraditional glycemic markers with clinical complications are largely uncharacterized. Large epidemiologic population-based cohort studies are needed to fully characterize long-term risk associations and to better establish the prognostic value of these biomarkers. Such studies would inform relevant clinical cut-points, performance of these markers for risk stratification, and comparative predictive ability. Second, clinical studies with repeat assessments of glucose and HbA1c and those involving continuous glucose monitoring studies are needed to rigorously characterize associations with average glucose in persons with type 1 and type 2 diabetes. Such studies may help establish construct validity and utility of nontraditional markers for monitoring glycemic control. Finally, randomized clinical trials can determine whether use of these tests can improve care and outcomes for persons with diabetes. It is possible that one or more of these biomarkers may be an efficient and appropriate alternative to HbA1c in some patients and strategies that combine multiple tests for glycemia may be beneficial in certain settings.