Abstract
Background
There is growing interest in fructosamine, glycated albumin, and 1,5-anhydroglucitol (1,5-AG) as alternative measures of hyperglycemia, particularly for use in settings where traditional measures (glucose and HbA1c) are problematic or where intermediate (2–4 week) glycemic control is of interest. However, reference intervals for these alternative biomarkers are not established.
Methods
We measured fructosamine, glycated albumin, and 1,5-AG in a community-based sample of U.S. black and white adults who participated in the Atherosclerosis Risk in Communities (ARIC) Study. We calculated reference intervals, evaluated demographic differences, and derived cutoffs aligned with current diagnostic cut-points for HbA1c and fasting glucose.
Results
In a healthy reference population of 1,799 individuals (mean age 55 years, 51% female, 15% black), the 2.5th and 97.5th percentiles, respectively, were 194.8 and 258.0 umol/L for fructosamine, 10.7 and 15.1 % for glycated albumin, and 8.4 and 28.7 ug/mL for 1,5-AG. Distributions differed by race, sex, and body mass index. Equivalent concentrations of fructosamine and glycated albumin corresponding to an HbA1c 6.5% (96.5 percentile) were 270.2 umol/L and 15.6 %, respectively. Equivalent concentrations of fructosamine and glycated albumin corresponding to a fasting glucose of 126 mg/dL (93.9 percentile) were 261.7 umol/L and 15.0 %, respectively.
Conclusions
The reference intervals for these biomarkers should inform their clinical use. Diagnostic cut-point equivalents for fructosamine and glycated albumin could be useful to identify persons with hyperglycemia in settings where fasting glucose or HbA1c are not available or where the interpretation of these traditional measures is problematic.
There is growing interest in non-traditional biomarkers of hyperglycemia, particularly for use in settings where intermediate (2–4week) glycemic control is of interest or where traditional measures (glucose and HbA1c) are problematic. For example, HbA1c may be influenced by alterations in erythrocyte lifespan or hemoglobin, independent of glycemia (1). Three molecules of particular interest are fructosamine, glycated albumin, and 1,5-anhydroglucitol (1,5-AG). These markers of chronic hyperglycemia are extracellular and therefore independent of changes in erythrocytes or hemoglobin.
Fructosamine reflects the binding of glucose to all serum proteins, predominately albumin but also other proteins including globulins and lipoproteins. Glycated albumin is a measure of glucose bound specifically to serum albumin and is commonly expressed as a percentage of total serum albumin. Due to the rate of turnover of serum proteins, fructosamine and glycated albumin concentrations correspond to ~2 to 4 weeks of past exposure to blood glucose. 1,5-AG is a monosaccharide that is normally stable in serum; it is ubiquitous in the diet. 1,5-AG closely resembles glucose in structure (it is the 1-deoxy form of glucose) and, like glucose, it is freely filtered by the glomeruli and normally reabsorbed in the renal proximal tubule. However, when glucose concentrations in the blood exceed the renal threshold (overt hyperglycemia, occurring at glucose concentrations of ~160 to 180 mg/dL), glucose will compete with 1,5-AG for reabsorption in the tubule, leading to urinary excretion of 1,5-AG and causing serum concentrations to drop. Thus, low serum 1,5-AG is thought to be a useful biomarker of hyperglycemic excursions occurring over the previous 1–2 weeks (2).
Reliable reference intervals are central to medical decision-making. A barrier to the routine clinical use of fructosamine, glycated albumin, and 1,5-AG is the lack of agreed-upon reference intervals or ‘normal’ values for these biomarkers. Previous studies examining the reference intervals and potential clinical cut-points for fructosamine or glycated albumin have been small (N<200) and the study populations have typically not been well characterized (3, 4). Furthermore, the literature on glycated albumin has largely focused on persons in Japan, Korea, and other Asian populations, where the assay is in wide clinical use (5–10).
The objective of this study was to define the reference intervals and demographic differences in fructosamine, glycated albumin, and 1,5-AG using data from a well-characterized community-based U.S. population of black and white adults. We also derived “HbA1c-equivalent” diabetes diagnostic cut-points for fructosamine and glycated albumin. The knowledge of clinically relevant cut-points for fructosamine or glycated albumin could be useful in settings where HbA1c is not available or in those conditions where its interpretation is problematic.
METHODS
Study population
We conducted this study using data from the community-based Atherosclerosis Risk in Communities (ARIC) Study, a large cohort of over 15,000 mostly black and white middle-aged adults from four U.S. communities: suburban Minneapolis, Forsyth County, North Carolina, Washington County, Maryland, and Jackson, Mississippi. Fructosamine, glycated albumin, and 1,5-AG data were available for participants who attended the second clinical examination, ARIC Visit 2, which took place from 1990 to 1992. To establish the reference intervals, we followed the approach established by the Clinical and Laboratory Standards Institute (11). To derive a “healthy” reference population, we included here participants without diagnosed diabetes. We excluded outliers using the Tukey approach (12) and sequentially excluded ARIC participants who were non-fasting, missing variables of interest, had serum album <3 g/dL, elevated liver enzymes, current smokers, had clinical or subclinical thyroid dysfunction, reduced kidney function, prevalent coronary heart disease, hypertension, or dyslipidemia. Our main analyses included a “healthy” reference population of 1,799 participants (eFigure 1). To derive cut-points for fructosamine and glycated albumin that were equivalent to diagnostic cut-points for HbA1c and fasting glucose, we used the larger ARIC Study population of 11,737 participants without a history of diagnosed diabetes (eFigure 1).
Study protocols were approved by institutional review boards at each study site and informed consent was obtained from all participants.
Laboratory Measurements
Measurements of fructosamine, glycated albumin, and 1,5-AG were conducted in 2012–2013 in stored serum samples originally collected from participants at Visit 2 (1990–1992) and stored at −70°C. These three assays were analyzed on the Roche Modular P800 analyzer at the University of Minnesota.
Fructosamine was quantitated using a colorimetric assay (Roche Diagnostics Corp.). The CVs were 3.2% at a concentration of 212.6 umol/L and 2.5% at a concentration of 856.7 umol/L%.
Glycated albumin was measured using an assay that requires separate measurements of total albumin (bromocresol purple) and glycated albumin (enzymatic method utilizing ketoamine oxidase and an albumin-specific protease) (Lucica GA-L Glycated Albumin, Asahi Kasei Pharma Corp.). The glycated albumin result was expressed as a percentage of total albumin using the manufacturer’s formula: [(glycated albumin concentration in g/dL/serum albumin concentration in g/dL)*100/1.14] + 2.9). The CVs for glycated albumin were 1.8% at a mean value of 56.0% and 2.1% at a mean value of 22.7%.
1,5-AG was measured using the GlycoMark assay (GlycoMark, Inc.). The CVs were 3.8% at a mean concentration of 4.6 μg/mL and 1.3% at a mean concentration of 14.7 μg/mL.
We compared the fructosamine, glycated albumin, and 1,5-AG results to fasting glucose and HbA1c measurements also available from participants at ARIC Visit 2. Serum glucose was measured as part of the original ARIC Study protocol using the hexokinase method (13). HbA1c was measured in EDTA whole blood using Tosoh instruments (Tosoh A1C 2.2 Plus Glycohemoglobin Analyzer in 2003–2004 and the Tosoh G7 method in 2007–2008, Tosoh Corp.) (14). Both Tosoh instruments are NGSP-certified, standardized to the Diabetes Control and Complications Trial assay, and performed in an NGSP Secondary Laboratory (University of Minnesota).
Statistical Analyses
We examined the distributions and summary statistics including the mean, standard deviation (SD), minimum, maximum, median, and percentiles (p2.5, p25, p50, p75, p95, p97.5) of fructosamine, glycated albumin, and 1,5-AG in our reference population. To evaluate the impact of the exclusions, we examined the change in the distribution after each exclusion. We derived reference intervals using a nonparametric approach based on the 2.5th and 97.5th percentiles and corresponding 90% confidence intervals in the reference population (15).
We visually displayed and compared the distributions across subgroups using forest plots and histograms. For fructosamine and glycated albumin only (since 1,5-AG is not proposed as a diagnostic test), we evaluated percentiles of each biomarker corresponding to diabetes diagnostic cut-points for HbA1c (5.7 and 6.5%) and fasting glucose (100 and 126 mg/dL) (16). We calculated the Pearson’s correlations between biomarkers and generated scatterplots with corresponding regression and lowess curves. We also specifically examined associations of the different biomarkers with body mass index. Because there is current debate regarding whether fructosamine should be corrected for total serum protein concentration, we compared the Pearson’s correlations of fructosamine with HbA1c and glycated albumin both before and after correction for serum albumin using published equations (17–20).
RESULTS
Characteristics of the referent population of 1,799 middle-aged adults are shown in Table 1. The means (SDs) for fructosamine, glycated albumin, and 1,5-AG were 225.8 (16.4) umol/L, 12.7 (1.1) %, and 18.4 (5.1) ug/mL, respectively (Table 2). By way of comparison, the means (SDs) for HbA1c and fasting glucose in this same reference population were 5.3 (0.3) % and 100.3 (9.3) mg/dL, respectively. The exclusion of diagnosed diabetes made a substantial impact on the distributions of the biomarkers, while the other exclusions had relatively small effects (eFigure 2). After exclusions, the reference intervals (2.5th and 97.5th percentiles) in the overall reference population were were: 194.8 and 258.0 umol/L for fructosamine; 10.7 and 15.1% for glycated albumin; and 8.4 and 28.7 ug/mL for 1,5-AG (Table 2).
Table 1.
Characteristics of referent population, healthy subsample of the Atherosclerosis Risk in Communities (ARIC) Study, n=1,799
| Variable | Mean (SD) or % |
|---|---|
| Age (years) | 55.3 (5.4) |
| Female, % | 51.4 |
| Black, % | 14.5 |
| Categories of body mass index | |
| <25 kg/m2 | 40.1 |
| 25 to <30 kg/m2 | 40.7 |
| ≥30 kg/m2 | 19.1 |
| LDL-cholesterol (mg/dL) | 104.6 (20.9) |
| HDL-cholesterol (mg/dL) | 15.0 (9.0) |
| Triglycerides (mg/dL) | 106.2 (53.1) |
| Systolic blood pressure (mmHg) | 113.2 (12.1) |
| Diastolic blood pressure (mmHg) | 69.8 (8.0) |
| eGFR (mL/min/1.73 m2) | 97.6 (12.4) |
Table 2.
Nonparametric reference intervals of fructosamine, glycated albumin, and 1,5-anhydroglucitol overall and according to sex, race, age, and body mass index, n=1,799
| n | Mean (SD) | 2.5th percentile (90% CI) | 25th percentile | Median | 75th percentile | 97.5th percentile (90% CI) | |
|---|---|---|---|---|---|---|---|
| Fructosamine, umol/L | |||||||
| Overall | 1799 | 225.8 (16.4) | 194.8 (192.1, 196.2) | 214.3 | 225.6 | 236.3 | 258.0 (256.4, 260.4) |
| Overall excluding BMI ≥30 kg/m2 | 1455 | 227.4 (16.1) | 197.4 (195.4, 198.9) | 216.5 | 226.9 | 238.0 | 259.2 (257.2, 264.0) |
| Sex | |||||||
| Male | 875 | 226.1 (15.9) | 193.7 (190.0, 196.6) | 216.0 | 226.4 | 236.6 | 256.5 (255.2, 258.0) |
| Female | 924 | 225.4 (16.9) | 195.4 (192.0, 196.7) | 213.4 | 224.9 | 235.8 | 263.3 (257.3, 267.7) |
| p-value* | 0.186 | 0.371 | |||||
| Race | |||||||
| Black | 261 | 231.8 (18.3) | 198.2 (195.4, 199.6) | 219.0 | 232.6 | 245.8 | 267.8 (263.8, 273.1) |
| White | 1538 | 224.8 (15.8) | 193.8 (191.8, 195.1) | 214.1 | 224.5 | 234.9 | 256.0 (254.5, 258.4) |
| p-value* | <0.001 | <0.001 | |||||
| Age (tertiles) | |||||||
| 47–52 years | 707 | 224.3 (16.2) | 191.9 (189.4, 195.9) | 213.2 | 224.8 | 235.1 | 254.5 (253.1, 258.0) |
| 53–57 years | 499 | 226.7 (16.9) | 194.8 (192.0, 197.4) | 215.2 | 225.7 | 237.7 | 265.1 (258.0, 268.2) |
| 58–68 years | 593 | 226.7 (16.1) | 196.5 (193.8, 198.3) | 215.2 | 226.2 | 237.7 | 258.7 (255.8, 263.4) |
| p-value* | 0.437 | 0.009 | |||||
| Categories of body mass index | |||||||
| <25 kg/m2 | 722 | 230.5 (16.1) | 200.0 (196.7, 202.5) | 220.3 | 230.3 | 241.2 | 263.6 (259.4, 266.8) |
| 25 to <30 kg/m2 | 733 | 224.3 (15.4) | 196.5 (193.8, 198.3) | 213.4 | 223.9 | 234.1 | 254.5 (252.1, 258.5) |
| ≥30 kg/m2 | 344 | 218.9 (16.1) | 188.3 (183.4, 190.8) | 208.1 | 217.8 | 229.3 | 251.7 (247.6, 256.1) |
| p-value* | <0.001 | <0.001 | |||||
| Glycated albumin, % | |||||||
| Overall | 1799 | 12.7 (1.1) | 10.7 (10.5, 10.8) | 11.9 | 12.7 | 13.5 | 15.1 (15.0, 15.3) |
| Overall excluding BMI ≥30 kg/m2 | 1455 | 12.9 (1.1) | 10.9 (10.8, 10.9) | 12.1 | 12.8 | 13.6 | 15.2 (15.1, 15.4) |
| Sex | |||||||
| Male | 875 | 12.6 (1.1) | 10.5 (10.3, 10.8) | 11.8 | 12.6 | 13.3 | 15.0 (14.7, 15.1) |
| Female | 924 | 12.9 (1.2) | 10.8 (10.6, 10.9) | 12.1 | 12.9 | 13.7 | 15.3 (15.1, 15.5) |
| p-value* | <0.001 | <0.001 | |||||
| Race | |||||||
| Black | 261 | 13.3 (1.2) | 10.9 (10.4, 11.3) | 12.6 | 13.3 | 14.2 | 15.5 (15.3, 15.6) |
| White | 1538 | 12.7 (1.1) | 10.7 (10.5, 10.8) | 11.9 | 12.6 | 13.4 | 14.9 (14.7, 15.1) |
| p-value* | <0.001 | <0.001 | |||||
| Age (tertiles, years) | |||||||
| 47–52 | 707 | 12.7 (1.1) | 10.8 (10.5, 10.8) | 11.9 | 12.6 | 13.4 | 14.9 (14.7, 15.4) |
| 53–57 | 499 | 12.8 (1.1) | 10.6 (10.4, 10.9) | 12.0 | 12.7 | 13.4 | 15.1(15.0, 15.3) |
| 58–68 | 593 | 12.8 (1.2) | 10.7 (10.4, 10.9) | 12.0 | 12.8 | 13.6 | 15.3 (15.1, 15.5) |
| p-value* | 0.087 | 0.079 | |||||
| Categories of body mass index | |||||||
| <25 kg/m2 | 722 | 13.1 (1.1) | 11.0 (10.9, 11.1) | 12.3 | 13.1 | 13.8 | 15.3 (15.2, 15.6) |
| 25 to <30 kg/m2 | 733 | 12.6 (1.1) | 10.8 (10.5, 10.9) | 11.9 | 12.6 | 13.3 | 15.0 (14.7, 15.2) |
| ≥30 kg/m2 | 344 | 12.3 (1.1) | 10.2 (10.1,10.3) | 11.5 | 12.3 | 13.0 | 14.4 (14.4, 14.8) |
| p-value* | <0.001 | <0.001 | |||||
| 1,5-anhydroglucitol, ug/mL | |||||||
| Overall | 1799 | 18.4 (5.1) | 8.4 (7.8, 8.8) | 15.0 | 18.3 | 21.8 | 28.7 (28.3, 29.1) |
| Overall excluding BMI ≥30 kg/m2 | 1455 | 18.3 (5.0) | 8.5 (8.1, 9.2) | 14.9 | 18.1 | 21.6 | 28.7 (28.3, 29.2) |
| Sex | |||||||
| Male | 875 | 19.6 (5.2) | 8.5 (7.4, 9.5) | 16.1 | 19.8 | 23.4 | 29.2 (28.7, 30.3) |
| Female | 924 | 17.3 (4.7) | 8.2 (7.7, 8.8) | 14.3 | 17.2 | 20.3 | 27.3 (26.5, 28.3) |
| p-value* | <0.001 | <0.001 | |||||
| Race | |||||||
| Black | 261 | 16.9 (4.7) | 7.7 (7.2, 8.6) | 14.0 | 16.4 | 20.0 | 27.2 (25.4, 28.7) |
| White | 1538 | 18.7 (5.1) | 8.5 (7.9, 9.1) | 15.2 | 18.6 | 22.1 | 28.8 (28.5, 29.5) |
| p-value* | <0.001 | <0.001 | |||||
| Age (tertiles, years) | |||||||
| 47–52 | 707 | 18.2 (5.1) | 8.3 (7.5, 8.8) | 14.7 | 18.1 | 21.6 | 28.6 (27.9, 30.2) |
| 53–57 | 499 | 18.7 (5.0) | 8.8 (7.5, 10.0) | 15.3 | 18.5 | 22.0 | 28.7 (27.6, 29.5) |
| 58–68 | 593 | 18.5 (5.1) | 8.0 (7.4, 9.2) | 15.2 | 18.3 | 21.7 | 28.7 (28.1, 29.5) |
| p-value* | 0.485 | 0.305 | |||||
| Categories of body mass index | |||||||
| <25 kg/m2 | 722 | 18.1 (5.0) | 8.9 (7.7, 9.7) | 14.6 | 17.7 | 21.4 | 28.7 (28.1, 29.9) |
| 25 to <30 kg/m2 | 733 | 18.5 (5.0) | 8.5 (7.9, 8.9) | 15.3 | 18.3 | 22.0 | 28.6 (27.8, 29.2) |
| ≥30 kg/m2 | 344 | 18.9 (5.2) | 7.5 (6.4, 8.4) | 15.6 | 19.3 | 22.3 | 28.9 (27.5, 30.2) |
| p-value* | <0.001 | 0.066 | |||||
p-value for difference in medians from nonparametric equality-of-medians test or difference in means from t-test (two groups) or ANOVA (three groups).
Concentrations of fructosamine were slightly higher in males compared to females, lower in whites compared to blacks, higher at older ages, and lower at higher categories of body mass index (Table 2 and eFigure 3, panel A). Patterns of demographic differences were similar for glycated albumin, except for sex where females had higher values compared to males (Table 2 and eFigure 3, Panel B). We observed inverse associations of fructosamine and glycated albumin with body mass index; this was in contrast to the positive associations of fasting glucose and HbA1c with body mass index (eFigure 4).
1,5-AG concentrations were lower in females compared to males, lower in whites compared to blacks, and lower at higher ages, but not significantly so (Table 2 and eFigure 3, Panel C). 1,5-AG was higher at higher categories of body mass index.
The concentrations of fructosamine corresponding prediabetes and diabetes clinical cut-points of 5.7% and 6.5% for HbA1c, based on percentiles, were 241.4 umol/L (the 77.1 percentile) and 270.2 umol/L (the 96.5 percentile), respectively (Table 3). The corresponding values of glycated albumin were 13.6% and 15.6%, respectively. The concentrations of fructosamine corresponding to prediabetes and diabetes clinical cut-points of 100 and 126 mg/dL for fasting glucose were 224.9 umol/L (the 45.3 percentile) and 261.7 umol/L (the 93.9 percentile), respectively. The corresponding values for glycated albumin were 12.5% and 15.0%, respectively.
Table 3.
Percentile equivalents for fructosamine, glycated albumin, and 1,5-anhydroglucitol based on clinical cut-points for HbA1c and fasting glucose in the total population of persons without diagnosed diabetes, n=11,737
| HbA1c | Fasting glucose | |||
|---|---|---|---|---|
| 5.7% | 6.5% | 100 mg/dL | 126 mg/dL | |
| Percentile: | 77.1 | 96.5 | 45.3 | 93.9 |
| Fructosamine, umol/L | 241.4 | 270.2 | 224.9 | 261.7 |
| Glycated albumin, % | 13.6 | 15.6 | 12.5 | 15.0 |
| 1,5-anhydroglucitol, ug/mL | 29.1 | 22.7 | 27.5 | 17.9 |
After correcting fructosamine for serum albumin using published equations, correlations of fructosamine with HbA1c and with glycated albumin were strengthened (eTable 2). Scatterplots and correlations of the different biomarkers are shown in eFigure 5.
DISCUSSION
The present study established reference intervals for three non-traditional assays of chronic hyperglycemia. The Roche fructosamine assay has been approved for clinical use for many years and is the dominate glycated protein assay used in the U.S. The Asahi Kasei Lucica GA-L glycated albumin assay is used widely in Japan and other countries in Asia and was FDA-cleared for clinical use in the U.S. in October 2017. The GlycoMark 1,5-AG assay is approved for clinical use in the U.S. and is reimbursed by Medicare and some other insurers, but is not widely used. Our results provide standard values that are likely to facilitate clinical interpretation of each of these assays.
The reference interval for fructosamine reported in the Roche package insert is 205 to 285 umol/L. This range was derived from data published in 1989 from 555 “apparently healthy” blood donors between 20 and 60 years of age (21). Our reference interval of 195 to 258 umol/L suggests that the normal range for this assay may need to be updated based on modern clinical performance data.
The distribution of glycated albumin in our study population was similar to a study of 1334 Italian blood donors which showed similar patterns for sex and age and a 97.5th percentile of 14.5% using an enzymatic assay (22). A previous study in 201 healthy U.S. subjects without known diabetes and normal glucose tolerance identified a reference interval of 11.9 to 15.8% for glycated albumin (4). Higher values in blacks compared to whites and females compared to males were also observed, although the sex difference was not statistically significant possibly owing to the limited power in this small study. The prediabetes diagnostic threshold for glycated albumin determined in the present study (13.6%) is similar to a previous cohort of 236 African immigrants which used the same glycated albumin assay and identified a threshold of 13.8% (the percentile equivalent of an HbA1c of 5.7% in their study population) (23).
Lower values of both fructosamine and glycated albumin at higher categories of body mass index have been previously reported (23–26), with U- or J-shape associations of fructosamine and glycated albumin with body mass index. This is in contrast to the associations of fasting glucose and HbA1c with body mass index, which tends to be positive and roughly linear. The reasons for lower concentrations of fructosamine and glycated albumin at high levels of adiposity remains unexplained but may relate to high levels of inflammation or issues related to protein turnover (27, 28).
The diabetes diagnostic cut-point of 6.5% for HbA1c was chosen for its specificity (29). Indeed, in our reference population, an HbA1c of 6.5% corresponded to the 96.5 percentile whereas the diagnostic fasting glucose cut-point of 126 mg/dL corresponded to the 93.9 percentile. Thus, the “equivalent” fructosamine and glycated albumin values were higher for HbA1c than fasting glucose. The “diagnostic” cut-point equivalents for fructosamine and glycated albumin provided in the present study may be useful in studies which do not have fasting blood samples for the measurement of glucose or whole blood samples for the measurement of HbA1c. Indeed, many cohorts have stored non-fasting plasma or serum in which fructosamine or glycated albumin could be reliably measured and used to determine the glycemic status of the study population.
Diagnostic cut-points are distinct from reference intervals and are typically derived based on a synthesis of multiple types of evidence and including, but not limited to, diagnostic testing studies, randomized clinical trials, epidemiologic evidence, and cost-effectiveness analyses. Recommendations for specific diagnostic cut-points are often highly political and controversial. Our goal here was not to debate the optimal diagnostic or screening cut-points for fructosamine or glycated albumin, but to equate these biomarkers to existing clinically relevant cut-points for HbA1c. The biology of fructosamine and glycated albumin are similar to HbA1c and correlations between these biomarkers in the setting of hyperglycemia are high (30, 31).
The 1,5-AG assay reflects hyperglycemia only when glucose exceeds the renal threshold. By design, our study population was limited to a “healthy” reference group. Thus, no participants in the present study had concentrations of blood glucose exceeding the renal threshold (maximum fasting glucose was 130 mg/dL in our study). 1,5-AG concentrations less than 10 ug/mL are believed to reflect “frequent” hyperglycemic excursions above the renal threshold (2). However, in our study population, there were 88 individuals (4.89%) with 1,5-AG concentrations less than 10 ug/mL. The demographic differences observed for 1,5-AG in the present study may reflect non-glycemic factors such as dietary differences or other determinants of 1,5-AG in persons without diabetes (32). Indeed, our recent genetic analyses suggest that 1,5-AG concentrations may also reflect the speed of glucose digestion and enteric uptake in persons without diabetes (33).
Some limitations of this study that should be considered in the interpretation of our results include that our study population was limited to middle-aged black and white adults (age range: 47 to 68; 15% black). Nonetheless, this population is likely largely generalizable to the majority of the U.S. population to whom these tests might be applied. For specific patient populations, such as pregnant women, additional studies are warranted. We did not have serum albumin measured with the bromocresol green assay method in this study. Corrections for fructosamine in the present study were conducted using the Asahi Kasei serum albumin (bromocresol purple). The assays examined in the present study are not formally standardized in the U.S. which suggests these results may not apply to other methods implemented at other labs. Nonetheless, because the Roche fructosamine assay is the predominant serum glycated protein assay in the U.S., our results for this assay may be fairly generalizable. Indeed, 95% of participants in the 2016 College of American Pathologists fructosamine (FT-B) survey used the Roche method and results are similar among laboratories (CV <3.5%), suggesting that standardization is not a substantial issue for this method in the U.S. (34). An additional limitation is that all measurements were conducted in long-term stored samples. Nonetheless, we demonstrated excellent analytical performance of these assays (CVs <4%) and prior studies have demonstrated high reliability of these assays in stored samples (35–39).
In conclusion, we defined a healthy population within the community-based ARIC cohort to establish reference intervals for fructosamine, glycated albumin, and 1,5-AG overall and in important demographic subgroups. We also identified fructosamine and glycated albumin cut-points corresponding to values of HbA1c and fasting glucose used for diagnosis and screening of diabetes. The results of this study should help inform the clinical use of the Roche fructosamine, Asahi Kasei glycated albumin, and GlycoMark 1,5-AG assays.
Supplementary Material
Acknowledgments
The Atherosclerosis Risk in Communities study has been funded in whole or in part with Federal funds from the National Heart, Lung, and Blood Institute (NHLBI), National Institutes of Health (NIH), Department of Health and Human Services, under Contract nos. (HHSN268201700001I, HHSN268201700003I, HHSN268201700005I, HHSN268201700004I, HHSN2682017000021). The authors thank the staff and participants of the ARIC study for their important contributions. This work was supported by NIH/NIDDK grant R01DK089174. Dr. Selvin was also supported by NIH/NIDDK grant K24DK106414. B. Warren was supported by NIH/NHLBI grant T32 HL007024. D.B.S. was supported by the Intramural Research Program of the National Institutes of Health Clinical Center. Reagents for the fructosamine assay were donated by the Roche Diagnostic Corporation. Reagents for the glycated albumin assays were donated by the Asahi Kasei Pharma Corp. Reagents for the 1,5-anhydroglucitol assays were donated by GlycoMark, Inc.
Footnotes
Supplemental material is provided in an Online Appendix.
Conflicts: The authors declare no conflicts of interest.
References
- 1.Welsh KJ, Kirkman MS, Sacks DB. Role of Glycated Proteins in the Diagnosis and Management of Diabetes: Research Gaps and Future Directions. Diabetes Care. 2016;39(8):1299–306. doi: 10.2337/dc15-2727. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dungan KM, Buse JB, Largay J, Kelly MM, Button EA, Kato S, et al. 1,5-anhydroglucitol and postprandial hyperglycemia as measured by continuous glucose monitoring system in moderately controlled patients with diabetes. Diabetes Care. 2006;29(6):1214–9. doi: 10.2337/dc06-1910. [DOI] [PubMed] [Google Scholar]
- 3.Melzi d’Eril GV, Bosoni T, Solerte SB, Fioravanti M, Ferrari E. Performance and clinical significance of the new fructosamine assay in diabetic patients. Wien Klin Wochenschr Suppl. 1990;180:60–3. discussion 78–81. [PubMed] [Google Scholar]
- 4.Kohzuma T, Yamamoto T, Uematsu Y, Shihabi ZK, Freedman BI. Basic performance of an enzymatic method for glycated albumin and reference range determination. J Diabetes Sci Technol. 2011;5(6):1455–62. doi: 10.1177/193229681100500619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Furusyo N, Koga T, Ai M, Otokozawa S, Kohzuma T, Ikezaki H, et al. Utility of glycated albumin for the diagnosis of diabetes mellitus in a Japanese population study: results from the Kyushu and Okinawa Population Study (KOPS) Diabetologia. 2011;54(12):3028–36. doi: 10.1007/s00125-011-2310-6. [DOI] [PubMed] [Google Scholar]
- 6.Hwang YC, Jung CH, Ahn HY, Jeon WS, Jin SM, Woo JT, et al. Optimal glycated albumin cutoff value to diagnose diabetes in Korean adults: a retrospective study based on the oral glucose tolerance test. Clin Chim Acta. 2014;437:1–5. doi: 10.1016/j.cca.2014.06.027. [DOI] [PubMed] [Google Scholar]
- 7.Li Q, Pan JM, Ma XJ, Bao YQ, Tang JL, Yuan QY, et al. Combined utility of hemoglobin A1c and glycated albumin in diabetic screening. Zhonghua Yi Xue Za Zhi. 2011;91(26):1813–6. [PubMed] [Google Scholar]
- 8.Ma XJ, Pan JM, Bao YQ, Zhou J, Tang JL, Li Q, et al. Combined assessment of glycated albumin and fasting plasma glucose improves the detection of diabetes in Chinese subjects. Clin Exp Pharmacol Physiol. 2010;37(10):974–9. doi: 10.1111/j.1440-1681.2010.05417.x. [DOI] [PubMed] [Google Scholar]
- 9.Ikezaki H, Furusyo N, Ihara T, Hayashi T, Ura K, Hiramine S, et al. Glycated albumin as a diagnostic tool for diabetes in a general Japanese population. Metabolism. 2015;64(6):698–705. doi: 10.1016/j.metabol.2015.03.003. [DOI] [PubMed] [Google Scholar]
- 10.Zhou Q, Shi DB, Lv LY. The establishment of biological reference intervals of nontraditional glycemic markers in a Chinese population. J Clin Lab Anal. 2016 doi: 10.1002/jcla.22097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Clinical and Laboratory Standards Institute (CLSI) Approved Guideline. Third. 30 Vol. 28. Clinical and Laboratory Standards Institute: 2010. Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory. [Google Scholar]
- 12.International Federation of Clinical Chemistry and Laboratory Medicine. Approved Guideline. Third. 30 Vol. 28. Clinical and Laboratory Standards Institute; 2010. Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory. [Google Scholar]
- 13.ARIC Protocol, Manual 10, Clinical Chemistry Determinations, Visit 2, Version 2.0. 1991 http://www.cscc.unc.edu/aric/visit/Clinical_Chemistry_Determinations.2_10.pdf. Last accessed: August 10, 2017.
- 14.Selvin E, Coresh J, Zhu H, Folsom A, Steffes MW. Measurement of HbA1c from stored whole blood samples in the Atherosclerosis Risk in Communities study. J Diabetes. 2010;2(2):118–24. doi: 10.1111/j.1753-0407.2010.00070.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Horowitz GL. Establishment and Use of Reference Values. 2012:95–118. [Google Scholar]
- 16.2. Classification and Diagnosis of Diabetes. Diabetes Care. 2017;40(Supplement 1):S11–S24. doi: 10.2337/dc17-S005. [DOI] [PubMed] [Google Scholar]
- 17.Howey JE, Browning MC, Fraser CG. Assay of serum fructosamine that minimizes standardization and matrix problems: use to assess components of biological variation. Clin Chem. 1987;33(2 Pt 1):269–72. [PubMed] [Google Scholar]
- 18.Lin MJ, Hoke C, Ettinger B, Coyne RV. Technical performance evaluation of BM/Hitachi 747–200 serum fructosamine assay. Clin Chem. 1996;42(2):244–8. [PubMed] [Google Scholar]
- 19.Mittman N, Desiraju B, Fazil I, Kapupara H, Chattopadhyay J, Jani CM, et al. Serum fructosamine versus glycosylated hemoglobin as an index of glycemic control, hospitalization, and infection in diabetic hemodialysis patients. Kidney Int Suppl. 2010;(117):S41–5. doi: 10.1038/ki.2010.193. [DOI] [PubMed] [Google Scholar]
- 20.Van Dieijen-Visser MP, Seynaeve C, Brombacher PJ. Influence of variations in albumin or total-protein concentration on serum fructosamine concentration. Clin Chem. 1986;32(8):1610. [PubMed] [Google Scholar]
- 21.Kruse-Jarres JD, Jarausch J, Lehmann P, Vogt BW, Rietz P. A new colorimetric method for the determination of fructosamine. LaboratoriumsMedizin/Journal of Laboratory Medicine. 1989:245. [Google Scholar]
- 22.Bellia C, Zaninotto M, Cosma C, Agnello L, Lo Sasso B, Bivona G, et al. Definition of the upper reference limit of glycated albumin in blood donors from Italy. Clin Chem Lab Med. 2017 doi: 10.1515/cclm-2017-0179. [DOI] [PubMed] [Google Scholar]
- 23.Sumner AE, Duong MT, Bingham BA, Aldana PC, Ricks M, Mabundo LS, et al. Glycated Albumin Identifies Prediabetes Not Detected by Hemoglobin A1c: The Africans in America Study. Clin Chem. 2016;62(11):1524–32. doi: 10.1373/clinchem.2016.261255. [DOI] [PubMed] [Google Scholar]
- 24.Poon AK, Juraschek SP, Ballantyne CM, Steffes MW, Selvin E. Comparative associations of diabetes risk factors with five measures of hyperglycemia. BMJ Open Diabetes Res Care. 2014;2(1):e000002. doi: 10.1136/bmjdrc-2013-000002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Miyashita Y, Nishimura R, Morimoto A, Matsudaira T, Sano H, Tajima N. Glycated albumin is low in obese, type 2 diabetic patients. Diabetes Res Clin Pract. 2007;78(1):51–5. doi: 10.1016/j.diabres.2007.02.021. [DOI] [PubMed] [Google Scholar]
- 26.Koga M, Matsumoto S, Saito H, Kasayama S. Body mass index negatively influences glycated albumin, but not glycated hemoglobin, in diabetic patients. Endocr J. 2006;53(3):387–91. doi: 10.1507/endocrj.k05-137. [DOI] [PubMed] [Google Scholar]
- 27.Koga M, Otsuki M, Matsumoto S, Saito H, Mukai M, Kasayama S. Negative association of obesity and its related chronic inflammation with serum glycated albumin but not glycated hemoglobin levels. Clin Chim Acta. 2007;378(1–2):48–52. doi: 10.1016/j.cca.2006.10.013. [DOI] [PubMed] [Google Scholar]
- 28.Chagnac A, Weinstein T, Herman M, Hirsh J, Gafter U, Ori Y. The effects of weight loss on renal function in patients with severe obesity. J Am Soc Nephrol. 2003;14(6):1480–6. doi: 10.1097/01.asn.0000068462.38661.89. [DOI] [PubMed] [Google Scholar]
- 29.The International Expert Committee. International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes. Diabetes care. 2009;32(7):1327–34. doi: 10.2337/dc09-9033. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Juraschek SP, Steffes MW, Selvin E. Associations of alternative markers of glycemia with hemoglobin A(1c) and fasting glucose. Clin Chem. 2012;58(12):1648–55. doi: 10.1373/clinchem.2012.188367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Selvin E, Rawlings AM, Grams M, Klein R, Sharrett AR, Steffes 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):279–88. doi: 10.1016/S2213-8587(13)70199-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Juraschek SP, Miller ER, 3rd, Appel LJ, Christenson RH, Sacks FM, Selvin E. Effects of dietary carbohydrate on 1,5-anhydroglucitol in a population without diabetes: results from the OmniCarb trial. Diabet Med. 2017;34(10):1407–13. doi: 10.1111/dme.13391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Li M, Maruthur NM, Loomis SJ, Pietzner M, North KE, Mei H, et al. Genome-wide association study of 1,5-anhydroglucitol identifies novel genetic loci linked to glucose metabolism. Sci Rep. 2017;7(1):2812. doi: 10.1038/s41598-017-02287-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.CAP surveys: participant summary for fructosamine survey 2016 set FT-B. Northfield, IL: College of American Pathologists; [Google Scholar]
- 35.Nathan DM, Steffes MW, Sun W, Rynders GP, Lachin JM. Determining stability of stored samples retrospectively: the validation of glycated albumin. Clin Chem. 2011;57(2):286–90. doi: 10.1373/clinchem.2010.150250. [DOI] [PubMed] [Google Scholar]
- 36.Selvin E, Rynders GP, Steffes MW. Comparison of two assays for serum 1,5-anhydroglucitol. Clin Chim Acta. 2011;412(9–10):793–5. doi: 10.1016/j.cca.2011.01.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Nathan DM, McGee P, Steffes MW, Lachin JM, Group DER Relationship of glycated albumin to blood glucose and HbA1c values and to retinopathy, nephropathy, and cardiovascular outcomes in the DCCT/EDIC study. Diabetes. 2014;63(1):282–90. doi: 10.2337/db13-0782. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Selvin E, Rawlings AM, Grams M, Klein R, Steffes M, Coresh J. Association of 1,5-anhydroglucitol with diabetes and microvascular conditions. Clin Chem. 2014;60(11):1409–18. doi: 10.1373/clinchem.2014.229427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Koskinen P, Irjala K. Stability of serum fructosamine during storage. Clin Chem. 1988;34(12):2545–6. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.