A pilotstudy: effect of erythrocyte lifespan determined by a modified…
A pilotstudy: effect of erythrocyte lifespan determined by a modified carbon monoxide breath test on glycosylated hemoglobin interpretation
Hong-Wei Chu 1, Yong-Jian Ma1-3G and Zhen-He Huang2
1 Guangdong Breath Test Engineering and Technology Research Center, Shenzhen University, Shenzhen, People’s Republic of China
2 Department of Endocrinology, Nanshan Affiliated Hospital of Shenzhen University, Shenzhen，Guangdong 518051,People’s Republic of China
3 Author to whom any correspondence should be addressed.
E-mail: mayj@szu.edu.cn
Keywords: carbon breath test, red blood cell lifespan, glycosylated hemoglobin, diabetes diagnostic criterion, correction
Abstract
Glycosylated hemoglobin (HbA1c) is an important criterion for the diagnosis of diabetes and an indicator of the blood glucose level. The red blood cell (RBC) lifespan heterogeneity is sufficient to influence the HbA1c interpretation. In this study, we recruited 120 patients with diabetes mellitus and 85 nondiabetic controls. The HbA1c and the RBC lifespan were detected by high-performance liquid chromatography and the advanced carbon monoxide breath detection method, respectively. Potential correlations of gender and age withHbA1c were analyzed and a receiver operator characteristic curve was generated to get the HbA1c cut-off for every RBC lifespan group. It was confirmed that HbA1c has no correlation with gender or age. The correlation formula between the HbA1c diagnostic criteria and RBC lifespan was derived to correct the HbA1c diagnostic criteria using the least-square method. The REC-lifespan-corrected HbA1c diagnostic criteria provided 100% sensitivity and specificity for the diagnosis of diabetes in the experimental set and was not refuted in the validated set. The diagnostic value of HbA1c is positively correlated with the RBC lifespan, and four patients with hyperglycemia, whose HbA1c values were lower than the general diagnosis criterion of 6.5%, were still considered to be diabetic according to this formula; that is, the application of this formula may help us to eliminate the 2.2% misdiagnosis rate of the current diagnostic criteria. To provide more accurate detection results, the effect of the RBC lifespan needs to be taken into account when HbA1c is used as a clinical indicator.
1.Introduction
Diabetes mellitus (DM) is a metabolic disease characterized by hyperglycemia. Long-term hyperglycemia in patients with diabetes can lead to chronic damage and dysfunction of various tissues, especially the eyes, kidneys, heart, blood vessels and nerves [1-5]. According to the 8th edition of IDF Diabetes Atlas, which was published by the International Diabetes Federation (IDF) in 2017, there are about 425 million adults 20-79 years of age with diabetes worldwide. The mortality rate of DM is as high as 70%, ranking third worldwide, second only to cardiovascular and cerebrovascular diseases and malignant tumors [6-10]. In China, which has 110 million diabetic patients, 300 billion yuan is spent annually on the treatment of DM and related diseases, accounting for approximately 13% of the national health care expenditure.
The common diagnostic indicators of DM are the fasting plasma glucose (FPG) level and 2-h plasma glucose value (2-h PG) during an oral glucose tolerance test (OGTT). The diagnostic indicators introduced by the American Diabetes Association (ADA) in 1997 are as follows: diabetic symptoms and a random plasma glucose > 11.1 mmoll-1 or an FPG >7.0 mmoll-1 or an OGTT 2-h PG > 11.1 mmoll-1. DM is diagnosed if any of the above criteria are met and one of the three criteria is met at a follow up visit the next day [5, 11 13]. This method is affected by diet, exercise and other factors, and only reflects the status that day, but does not reflect the long-term blood glucose level.
Glycosylated hemoglobin (HbA1c) is a product of hemoglobin in red blood cells (RBC) and glucose in the serum, so HbA1c can reflect the blood glucose level over the last 2-3 months and is not affected by recent changes in diet or activity [14, 15]. Selvin et al found that 2-h PG levels [within-person coefficient of variation (CV), 16.7%; 95% CI: 15.0-18.3] and FPG (CV,5.7%; 95% CI: 5.3-6.1) had substantially more variability compared with HbA1c (CV, 3.6%; 95% CI: 3.2-4.0) levels, which suggested that HbA1c has a greater analytic stability in comparison to the other two indicators [13, 16]. In 2010, the ADA added HbA1c (>6.5%) to the diagnostic criteria for DM. The World Health Organization (WHO) also recommended that HbA1c be used to diagnose DM in countries and regions where available [12, 13, 17 19].
Some studies suggest that for every 2 mmoll-1 increase in average blood glucose level, HbA1c increases by 1% [20]. The relationship between HbA1c and the average blood glucose level is that the average blood glucose (mmoll-1) = 1.98 × HbA1c-14.29 [21]. How ever, recent studies have shown that the HbA1c level is not completely consistent with the average blood glucose level [22］. When HbA1c is 6.0%, the corresponding average blood glucose level is 100 152 mg dl-1, while when HbA1c is 7.0%, the corresponding average blood glucose level is 123 185 mg dl-1. There is considerable overlap between the average blood glucose level of the two HbA1c values, and the variation in average blood glucose levels is beyond the normal range of fluctuation [23]. Kramer et al reported that the specificity of the HbA1c cut-off of 6.5% can reach 79% based on the Rancho Bernardo study; however, the sensitivity is only 44%, which would be further from providing an accurate diagnosis [17，24]. HbA1c is not only related to the blood glucose level, it is also closely related to the contact time between hemoglobin and blood glucose.τ'he index reflecting this contact time is the RBC lifespan [25]. Previous studies have suggested that for every 2 mmoll-1 increase in average blood glucose level, HbA1c increases by 1%, but this premise assumes that the RBC lifespan of patients with diabetes is the same. The lifespan of HbA1 c is consistent with that of the RBC. The longer the RBC lifespan, the greater the longest reaction time between hemoglobin and blood glucose. The RBC lifespan varies among individuals [26-29]. Therefore, if the RBC life span is not considered and an HBA1c value of 6.5% is used as the diagnostic criterion for diabetes, false negative or -positive results may occur, which means that it is necessary and indispensable to correct the diagnostic value of HbA1c by the RBC lifespan.
As far as we know, no credible corrections for the diagnostic value of HbA1c have been completed by any institution, which is mainly due to the complexity and difficulty in determining the RBC lifespan. Cohen et al used ex vivo biotin labeling and flow cytometric quantitation to determine the survival of autologous RBCs [30]. Khera et al measured the RBC lifespan in 10 adults by using biotin oral stable isotope labeling and confirmed that the variation in RBC lifespan could influence the HbA1c interpretation [31]. But these methods take too long to complete the statistical analysis of large samples. We used the advanced car bon monoxide (CO) breath test, which measures the CO content decomposed from hemoglobin to estimate the RBC lifespan, and can complete the determination of large samples to get the critical value of HbA1c corresponding to the RBC lifespan and the effect of RBC lifespan on HbA1c for providing a more accurate judgment .
2. Research design and methods
2.1.Subjects
120 subjects with type 2 DM were recruited in this study. The ratio of males to females was 7:5. DM was diagnosed by blood glucose criteria in all subjects. Patients who met the following conditions were excluded: baseline serum creatinine > 1.5 mg dl-1; urine albumin >200 µg min 1; hematocrit <34%; reticulocyte count > 2%; history of gastrointestinal blood loss; pregnancy; underlying illness known to be associated with body wasting (e.g. malignancies or tuberculosis); and a history of smoking. At the same time, 85 non-diabetic (NDM) controls were recruited, with a male-to-female ratio of 46:39. None of the subjects had apparent physiologic abnormalities and were strictly diagnosed without DM based on blood glucose criteria; the exclusion criteria were the same as for the DM subjects. The study protocol was approved by the Ethics Evaluation Committee of Shenzhen Sixth People’s Hospital. The experiments were conducted in accordance with the Helsinki Declaration. Before participation in the study, written informed consent was obtained from each subject.
2.2. Experimental design
All Subject, including DM and NDM subject，were required to provide peripheral venous blood samples at the Sixth People’s Hospital of Shenzhen for hemoglobin and HbA1C determinations based on high performance liquid chromatography. The breath samples were collected on the same day, and the samples were used to measure the RBC lifespan using an ELS TESTER™ (Seekya Biotec Co. Ltd, Shenzhen, China) at the Guangdong Breath Test Engineering and Technology Research Center. The breath samples were collected at 8-11am without fasting. Each participant held his or her breath for 10 s, then exhaled into a sample collection system through a mouthpiece （figure 1). As shown in figure 1, the collection system discards the first 300 ml of dead space gas, then directs subsequent alveolar air automatically into a foil collection bag. If needed, the procedure was repeated until the collected air sample reached the collection bag capacity of 1000 ml. The filled bag was detached and sealed. Atmospheric samples were collected just after breath sampling. Alveolar air and atmospheric samples were stored at room temperature and analyzed within 5 days. Most of the samples were used to explore the correlation between HbA1c diagnostic point and RBC lifespan, and a small remaining sample verification set was used to verify the accuracy of the results.
2.3. Instruments and detection
The Guangdong Breath Test Engineering and Tech nology Research Center has the newest automatic instrument for RBC lifespan determination (ELS TESTER™), which determines the alveolar endogenous CO concentration by non dispersive infrared spectroscopy with paired alveolar and air gas samples, and uses that measurement as a basis for determining the RBC lifespan with Levitt’s formula [28, 32-36]. The instrument design is shown in figure 2. The detection limit of the instrument for CO is 150 ppb, with an accuracy of ± 50 ppb and a precision within SO ppb. The instrument is easy to operate. A series of automatic measurements can be completed within 20 min by simply connecting the gas sample to the entrance, inputting the hemoglobin concentration data of the subjects, and pressing the start button. The detailed steps are as follows. First, the quality of alveolar samples was checked by measuring the concentration of C02 , which was used as the dilution indicator of alveolar samples. A small amount of alveolar air was pumped into the C02 detection chamber to measure the digital voltage under infrared (wavelength = 4.26 µm). The digital voltage was converted to C02 concentration by the pre-input calibration curve obtained from the standard C02 gas sample (voltage/C02 concentration curve). Only alveolar gas samples containing <5% C02 were considered diluted. Second, to eliminate the molecule interfering with CO infrared detection, the sample gas was pushed into the de interference system with an absorbent mixture consisting mainly of soda and asbestos, and the absorbed interfering molecutle was mainly water and C02 . Third, the difference in CO concentration between alveoli and atmosphere was measured using a paired measurement technique. The two de-interfering samples were pumped into the CO detection chamber in series, and the digital voltage was measured under infrared light (wavelength = 4.65 µm). The digital voltage difference between the paired samples was calculated by subtraction. Then, the digital voltage difference was converted to a CO concentration difference, which was recorded as the endogenous alveolar CO concentration via a pre-imputed calibration curve obtained with a series of paired standard CO gas samples of known concentrations (voltage difference/CO concentration difference curve). Fourth, any endogenous alveolar samples found to be diluted (i.e. <5% C02) were nomalized to a 5% C02 status. Finally, the RBC lifespan was calculated using the Levitt formula (see below) based on the pre-input hemoglobin concentration and the parameters corrected (if necessary) for endogenous alveolar CO concentration. The instrument reported the following data: alveolar C02 , endogenous alveolar CO and RBC lifespan [32].
2.4. CO exhalation method
There are three sources of CO in human breath: hemoglobin degradation, non hemoglobin degradation and exogenous inhalation. The proportion of hemoglobin degradation reaches 86% while 85% of this hemoglobin comes from RBC. Therefore, approximately 70% of endogenous CO in breath comes from RBC degradation. On the premise of excluding external interference, the rate of CO excretion in alveoli can be used to estimate the rate of RBC destruction [32, 37], and the time required for total hemoglobin decomposition is the RBC lifespan. Based on these principles, Levitt’s team developed a simple non invasive breathing test in 1992 for estimating the RBC lifespan in human subjects. In 2003, the same team modified the test, and the endogenous composition of CO in alveolar respiration was determined by subtracting the partial pressure of CO in atmosphere from the partial pressure of CO in alveolar air [26]. The RBC lifespan (in days) is calculated from CO exhalation according to the following formula, which equates the mean RBC lifespan to the total capacity of CO from hemoglobin divided by the CO quantity released per day:
where Vb is the blood volume while Vtis the volume of the resting alveolar ventilation which is equal to Vb in value, so this equation can be simplified as follows:
2.5. Statistical analysis
The subjects were grouped according the RBC lifespan which was slightly lower in DM subjects than in NDM subjects (p < 0.05). Subjects in the same age range were grouped into the same lifespan group. The specific grouping method was as follows: 50 days was taken as the lower limit for the first group; the lower limit plus 10% error was taken to obtain the midpoint value of the first group; the midpoint value of the first group plus 10% error was taken to obtain the upper limit value of the first group; one day was added to the upper limit of the first group as the lower limit value of the second group; and multiple lifespan groups were obtained by this analogy. RBC lifespan was given as the means with SDs, the difference between HbA1c and gender was assessed using t-tests, Pearson and Spear man analyses were used to analyze correlations between age and HbA1c, and the receiver operator characteristic (ROC) curve was obtained to get the diagnostic points of HbA1c in each group. The final diagnostic point was the HbA1c value when the combination of sensitivity and specificity reached the optimal value. The statistical analyses were conducted in SPSS 22.0 (SPSS for Windows, version 22, Chicago, IL, USA).p < 0.05 was considered significant.
2.6. Data and resource availability
(1) Data availability statements
. The data sets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.
(2) Resource availability statements. No applicable resources were generated or analyzed during the current study.
3.Results
Four NDM and five DM subjects were excluded from the analysis due to irregular measurement behavior. The remaining subjects had a wider distribution of RBC lifespan, which was slightly lower in DM subjects than in NDM subjects (p < 0.05). The RBC lifespan was 105.5 ± 28.5 days in NDM subjects versus 86.6 ± 23.0 days in DM subjects. The HbA1c in NDM subjects was 5.3 ± 0.4 (%), while in DM subjects it was 10.1 ± 2.3 (%), and they overlapped at the value of 5.8%. According to the grouping criteria, all the data were grouped into six lifespan groups. The data statistics are shown in table 1.
Specific HbA1lc data for the six groups of subjects are shown in figure 3.The sample amounts of the first and last groups were too small to be included in the follow-up statistical analysis. The HbA1c values of males and females were 8.2% ± 3.1 and 7.8% ± 2.8, respectively. There was no significant difference between gender and HbA1c (p > 0.05). In addition, HbA1c was not correlated to age while there was no significant difference between them (p > 0.05). For each group of DM and NDM subjects, the best critical value was found by SPSS 22.0 statistical analysis. As shown in table 2, the sensitivity and specificity of the critical value of HbA1c in each lifespan group reached 100%. It can be seen from the table that the critical value of HbA1c was positively correlated with the RBC lifespan. We plotted the curve with the midpoint of each RBC lifespan group as the horizontal ordinate and the critical value of HbA1c of each group as the longitudinal coordinate, as shown in figure 4. Through the least square method, the relationship between the critical value of HbA1c and RBC lifespan was obtained as follows. The fitting degree was 96.74%:
Critical value of HbA1c(%)
= 0.0151 * RBC lifespan (day) + 4.6451,
the critical value of HbA1c is the corresponding diagnostic point of the group with the RBC lifespan value, and HbA1c measured(HbA1cmeasure) value could be corrected to correspond to the average RBC lifespan of human as the HbA1c diagnostic point(HbA1cdi gnose) is 6.5% at the average RBC life span, the corrected value(HbA1ccorrect) is,
HbA1ccorrect = HbA1Cmeasure – (HbA1cdiagnose - 6.5)
= HbA1cmeasure - 0.0151
* RBC lifespan + 1.8549
ten DM subjects (five men and five women) and ten NDM subjects （five men and five women) were included in the validation set to verify the correctness of the formula. All subjects’ HbA1c diagnostic points were obtained according to their RBC lifespan, and no difference was found between the diagnosis based on blood glucose and HbA1c diagnostic point.
4. Discussion
In this study, 205 subjects (including DM and NDM subjects) were recruited through strict screening criteria. The value of HbA1c and the hemoglobin concentration were determined in the same hospital using the same method. The use of a highly efficient automated RBC lifespan-measuring instrument provided the benefit of enabling this larger sample size of participants relative to previous studies analyzing the RBC lifespan. The lifespan was 105.5 ± 28.5 days in NDM subject and 86.6 ± 23.0 days in DM subject , which was almost consistent with the experimental conclusion of Huang et al (86.1 ± 18.1 days) [38). Each subject was grouped according to the RBC lifespan value. We obtained the critical value of HbA1c corresponding to each lifespan group through data analysis and the relationship between the critical value of HbA1c and the RBC lifespan through the least-square method. The sensitivity and specificity of the critical value of HbA1c for DM in each lifespan group reached 100%, so there is reason to believe that the critical HbA1c obtained can be used as the cut off for the diagnosis of DM. From the formula between HbA1c and RBC lifespan we can see that with the increase in the lifespan of RBC, the diagnostic cut off for HbA1c is not a fixed value, but will increase. Therefore, when we use the HbA1c to diagnose DM, it is necessary to make adjustments according to the RBC lifespan.
HbA1c is defined as the stable adduct of glucose at the N terminal amino group of the β chain of hemoglobin Ao (N [1-deoxyfructosyl] hemoglobin) [15], which is an index reflecting the average blood glucose level in the past 2-3 months. HbA1c can be used to diagnose DM in the test subjects and is also the accepted measure for evaluating long-term blood glucose control. Compared with the blood glucose index, diagnosis with HbA1c does not require a fasting blood sample, the result is not affected by meals, it can reflect the long-term blood glucose level, it is less affected by short-term diet, exercise and other lifestyle changes, and errors caused by other non-glycemic factors affection HbA1c are rare, such as hemoglobinopathy [39].
However, the diagnostic value of HbA1c for DM has not been accurately and scientifically determined. In 2009, an international expert committee consisting of the ADA, the European Association for the study of Diabetes and the IDF issued a report on the application of HbA1c in the diagnosis of DM on the basis of reviewing three cross-sectional epidemiologic data sets from health and nutrition surveys in Egypt, the Pima Indians and the United States [40]. The international expert committee pointed out that an HbA1c of 6.5% was used to diagnose DM, which has sufficient sensitivity and specificity. In 2010, the ADA added an HbA1c value >6.5% in the diagnostic criteria of DM. Zemlin et al studied 946 DM subjects and NDM subjects and found that only half the patients with diabetes using the fasting glucose were regarded so with the 6.5% HbA1c criterion, whereas the sensitivity and specificity were 80% and 77% with 6.1% HbA1c [41]. In Chinese adults, the best diagnostic cut-off HbA1c level for DM is 6.2% 6.4%, with 6.3% comprising the majority [42, 43]. The uncertainty of the diagnostic cut-off for HbA1c is related to the degree of standardization of the detection method [25] and does not exclude ethnic differences, but the variation in the lifespan of individual RBCs may have an even greater impact on it.
In our study, we excluded those who might have possible modified HbA1c values due to e.g. pregnancy or hyperthyroidism when recruited. All the recruited subjects were assessed in the same hospital to exclude the interference of detection methods. Furthermore, the subjects had the same nationality and did not meet any weak socio-economic and psychological criteria. It is shown that gender and age are not correlated with HbA1c by statistical analysis (p < 0.05). The HbA1c value used clinically to diagnose diabetes and evaluate glycemic control depends on three main factors:(1) the HbA1c in reticulocytes when they are released from the bone marrow; (2) the synthetic rate of HbA1c (or Hb glycation rate) as RBCs become older; and(3) the mean age of RBCs in the circulation [30]. Blood HbA1c is a mean value for all RBCs with values that range from very low for reticulocytes to very high for the oldest. Nuttall et al found no significant correlation between glycosylated hemoglobin and reticulocyte count, and it is impossible that variation in reticulocyte HbA1c has a significant impact on HbA1c interpretation because of its low value, so when con sidering the HbA1c interpretation we can ignore the reticulocyte HbA1c [44]. The synthetic rate of HbA1c should be a function of glucose concentration to which hemoglobin is exposed and be higher in DM compared with NDM subjects [30]. The RBC lifespan is the longest reaction time hemoglobin is exposed to blood glucose and approximately twice the mean age of RBCs, so the effect of the RBC lifespan on the HbA1c value should not be ignored.
The most advanced CO method and the highly efficient, automated RBC lifespan-measuring instrument make measuring a large number of RBC life spans and exploring the correlation with HbA1c a reality. We found that the HbA1c values of DM and NDM subjects overlapped at the value of 5.8%, which probably means that the effect of HbA1c in the diagnosis of DM may disappear without considering the RBC lifespan. The formula between HbA1c and RBC lifespan was obtained, which reminds us that it cannot be ignored when HbA1c plays a role, and it has not been refuted by the validation set. For example, 6.1% should be selected as the cut off for the diagnosis of DM according to the relationship between the cut off for HbA1c and the RBC lifespan for a man whose RBC lifespan is 100 days; however, if we do not consider the influence of the RBC lifespan and still use 6.5% as the cut-off，the original patients with DM may be mis diagnosed as healthy people, resulting in false-negative results. Equally, when assessing blood glucose levels for patients, it may also lead to inappropriate judgments. Interestingly, three patients were diagnosed with diabetes whose FPG values were 7.5, 9.4, 10.4 and 7.8, respectively; however, the HbA1c values were lower than the general diagnosis criterion of 6.5% (5.8, 5.8, 6.1 and 6.4, respectively), but the modified HbA1c diagnostic points (5.78, 5.73, 5.99 and 6.25) told us they were really at risk of hyperglycemia. That means that if we confirm that blood glucose is the ideal indicator to judge whether a patient has diabetes or not, there will be about a 2.2% probability of misdiagnosis in the current diagnosis standard of HbA1c (6.5%). Similarly, there will be a certain degree of error when using HbA1c to evaluate the blood glucose level.
For the convenience of clinical application, we can leave the existing diagnostic point (6.5%) unchanged, but correct the HbA1c to the corresponding value of the human average RBC lifespan according to the adjusted diagnostic point, and make the diagnosis of DM through the 6.5% diagnostic point corresponding to the human average RBC lifespan.
This current study did not consider factors such as race or altitude due to the limitation of time and resources, and these related studies are being carried out with the expansion of the sample, which will be detailed in a follow-up study.
5. Conclusion
HbA1c is an important index for the diagnosis of DM and determination of the blood glucose level. The RBC lifespan will affect the HbA1c interpretation to some extent. In the past, it was difficult to explore whether HbA1c was affected by RBC lifespan because of the inconvenience in detecting the RBC lifespan. In this study, the most advanced CO breath detection method was used to help us to realize large-sample exploration, and the approximate relationship between the HbA1c value and RBC lifespan was obtained through statistical analysis. Based on the formula, it can be seen that the diagnostic cut-off for HbA1c was positively correlated with the RBC lifespan. The application of this formula may help us to eliminate at least 2.2% misdiagnosis rate and deepen our understanding of a patient’s blood glucose status. In order to make a more accurate move, the RBC lifespan should be considered when providing diagnosis and blood glucose assessment for test subjects.
Acknowledgments
The authors thank Zhen He Huang for his help in academic theory, and Jun Feng Luo and Cong Zhou for their help in the research process.
Funding
This study was supported by the Guangdong Breath Test Engineering and Technology Research Center,
and the detection instrument (ELS TESTER™) was provided by Seekya Biotec Co. Ltd, Shenzhen, China.
Author contributions
Hong-Wei Chu designed the study, acquired, analyzed and interpreted the data, wrote the initial draft of the manuscript, revised the manuscript to its final form, and is the guarantor of this work. As such, Chu had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Yong-Jian Ma researched the data, contributed to the study design and interpreted the data. Zhen He Huang contributed to the collection and measurement of the data.
Duality of interest
No potential conflicts of interest relevant to this article were reported .
ORCID iDs
Yong-Jian Ma O https://orcid.org/0000-0002- 4932-4562
References
[1] McEvoyP l974 Diabetes mellitus NursingTimes 701074-5
[2] Kuller LH etal 2000 Diabetes mellitusArter. Thromb. Vase. Biol.20 823
[3] Pedicino D etal 2012 Type 2 diabetes, immunity and cardiovascular risk: a complex relationship Pathophysiology and Complications of DiabetesMellitus (Rijeka: IntechOpen) pp 45-70
[4] Aok I，Shimoyama K, Aoki N, Homori M, Yanagisawa A, Nakaha ra K, Kawai Y, Kitamura SI and Ishikawa K 1996 Platelet-dependent thrombin generation in patients with diabetes mellitus: effects of glycemiccontrol on coagulability in diabetesJ.Am. Coll.Cardiol. 27560 6
[5] Malchoff CD 2012 Diagnosis and classification of diabetes mellitus Connecticut Medicine 35 564-71
[6] NCD Risk Factor Collaboration (NCD-RisC) 2016 Worldwide trends in diabetes since 1980: a pooled analysisof 751 population based studieswith 4*4 million participants Lancet 387 1513-30
[7] Cho NH, Shaw TE, Karuranga S, Huang Y, da Rocha Fernandes T D,Ohlrogge AWand Malanda B 2018IDF diabetes atlas: global estimates of diabetes prevalence for 2017 and projections for 2045 Diabetes Res.Clin. Pract. 138271-81
[8] IDF Diabetes Atlas Group 20 15 Update of mortality attributable to diabetes for the IDF DiabetesAtlas: estimates fortheyear 2013 Diabetes Res. Clin.Pract. 109 461-5
[9] Int. Diabetes Federation 2000 TDF Diabetes Atlas lstedn (Brussels: Int. Diabetes Federation) Beagley T, Guariguata L, Weil Cand Motala AA 20 14 Global estimates of undiagnosed diabetes in adultsDiabetes Res Clin. Pract. 103150 60
[10] International Diabetes Federation 2015IDF Diabetes Atlas 7th edn (Brussels:Int. Diabetes Federation)
[11] Eriksson T, Qiao Q and Tuomilehto J 1998 Will new diagnostic criteria for diabetes mellituschange phenotype of patients with diabetes? Reanalysis ofEuropean epidemiological data Br. Med.J. 317 371-5
[12] Vegt FD et al 1998 The 1997 American Diabetes Association Criteria Versus the 1985 World HeaIth Organization Criteria for the Diagnosis of Abnormal Glucose Tolerance: poor agreement in the hoorn study Diabetes Care 21 1686--90
[13] Kumar R et al 2016 Evidence for current diagnostic criteria of diabetes mellitus World J. Diabetes 7 396-405
[14] Kuenen J, Borg R, Zheng H, Schoenfeld Dand Heine R J 2008 Translatingthe A lC assay into estimated average glucose values Diabetes Care311473-8
[15] Jeppsson J O et al 2002 Approved !FCC reference method for the measurement of HbAlcin human blood Clin. Chem.Lab. Med. 40 78-89
[16] Selvin E 2007 Short-term variability in measures of glycemia and implications for the classifcation of diabetes Arch.Intern. Med. 167 1545
[17] Saudek C D, Derr R Land Kalvani R R 2006 Assessmg glycaemia in diabetes usingself-monitoring blood glucose and hemoglobin AlcJAMA 2951688 97
[18] American Diabetes Association 2007 Standardsof medical care of diabetes Diabet, Care 3054-41
[9] Goldstein DE, Little R,Lorenz RA, Malone J I, Nathan DM, Peterson C M and Sacks D B 2004 Tests of glycaemia in diabetes Diabetes Care 27 1761-73
[20] Thevarajah M, Nadzimah M N and Chew YY 2009 Interference of hemoglobinAlc( HbA1c) detection using ion exchange high performance liquid chromatography(HPLC) method by clinically silent hemoglobin variant in University Malta Medical centre(UMMC) -a case report Clin.Biochem. 42 430-4
[21] Rohlfing C L et al 2002 Defining the relationship between plasma glucose and HbA(1c):analysis of glucose profiles and HbA(1c)in the diabetes control and complications trial Diabetes Care 25 275-8
[22] Christensen DL and Witte D R 2010 Moving to an AIC-based diagnosis of diabetes has a different impact on prevalence in different ethnic groups Diabetes Care 33580-2
[23] Nathan D M et al 2013 Diabetes control and complications trial/epidemiology of diabetes interventions and complications study at 30 years: advances and contribut10ns Diabetes.62 3976-86
[24] Kramer CK, Araneta MR G and Barrett Connor E 2010 Al C and diabetes diagnosis: the rancho bernardo study Diabet， Care 33 l0l-3
[25] Goldstein DE et al 1995Tests of glycemia in diabetes Diabetes Care 18 896 909
[26] Berlin N I,Waldmann TA and Weissman SM 1959 Lifespan of red blood cell Physiol. Rev. 39 577-616
[27] Eadie G Sand BrownI W Jr 1955 The potential life span and ultimate survival of fresh red blood cells in normal healthy recipients as studied by simultaneous Cr5l tagging and differential hemolysis J. Clin. Invest. 34 629-36
[28] Strocchi A et al 1992 A simplecarbon monoxide breath test to estimate erythrocyte turnover J. Lab. Clin. Med. 120392 9
[29] Shemin D and Rittenberg D 1946Thelife span of the human red blood cell J. Biol. Chem. 166 627
[30] Cohen R M, Franco RS, Khera P K,Smith EP, Lindsell C J, Ciraolo P J, Palascak M B and Joiner C H 2008 Red cell lifespan heterogeneity in hematologically normal people is sufficient to alter HbA1c.Blood 1124284-91
[31] Khera PK et al 2015 Use of an oral stable isotope label to confirm variation in red blood cell mean age that influences HbA1c interpretation Am. J. Hematol. 90 50 5
[32] ZhangH-D et al 20 18 Human erythrocyte lifespan measured by Levitt's CO breath test with newly developed automatic instrument J. Breath Res 12 036003
[33] Strocchi A, Schwartz S, Ellefson M, Engel RR, Medina A and Levitt MD 1992A simple carbon monoxide breath test to estimate erythrocyte turnover J. Lab. Clin.Med. 120392 9
[34] Fume J K, Springfifield J R,Ho S B and Levitt MD 2003 Simplifification of the end-alveolar carbon monoxide technique to assess erythrocyte survival J. Lab. Clin.M ed. 142 52-7
[35] Ma Y J et al 2016 A modifified carbon monoxide breath testfor measuring erythrocyte lifespan in small animals Biomed. Res Int. 2016 7173156
[36] Ma Y J, Zhang H D, Wu CH,Zhu G L, Ji Y Q, Huang J L,Du L T, Cao P, Zang DYand Ji K M 20 16 Rapid CO breath test screening of drugs for protective effects on ribavirin-induced hemolysis in a rabbit model: a pilot study J. Breath Res. 10 0360 10
[37] Vreman H J, Wong R J and Dir S 2000 Carbon Monoxide in Breath, Blood, and Other Tissues Carbon Monoxide Toxicity 1 19-60
[38] Huang Z et al 2017 Relationship between glycated haemoglobin concentration and erythrocyte survival in type 2 diabetes mellitus determined by a modified carbon monoxide breath test J. Breath Res. 12026004
[39] UK Prospective Diabetes Study (UKPDS) Group 1998 Intensive blood- glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33) Lancet 352 837-53
[40] Nathan B et al 2009 International expert committee report on the role of the Al C assay in the diagnosis of diabetes Clin. Biochem. Rev. 32 1327-34
[4 1] Zemlin A E et al 2011 HbA1c of 6.5% to diagnose diabetes mellitus-does it work for Us？- the bellville south africa study PLoS One 6 e22558
[42] Bao Y,MaX, Li H, Zhou M, Hu C, Wu H,Tang J,Hou X, Xiang Kand Jia W 2010Glycated haemoglobin Ale for diagnosing diabetesin Chinese population: cross sectional epidemiological survey Br.Med.J. 340 1178
[43] Yu Y et al 2012 Validity of glycated hemoglobin in screening and diagnosing Type 2 diabetes mellitusin chinese subjects Korean J. Intern.Med.27 41 6
[44] Nuttall F Q et al 2004 Stability over time of glycohemoglobin, glucose, and red blood cell survival in hematologically stable people without diabetes Metabolism 531399 404
