Hba1c and Glycated Albumin Levels Are High in Gastrectomized Subjects with Iron-Deficiency Anemia
Total Page:16
File Type:pdf, Size:1020Kb
Available online at www.annclinlabsci.org 52 Annals of Clinical & Laboratory Science, vol. 47, no. 1, 2017 HbA1c and Glycated Albumin Levels Are High in Gastrectomized Subjects with Iron-Deficiency Anemia Shinya Inada1 and Masafumi Koga2 1Department of Internal Medicine, Kawanishi City Hospital and 2Department of Internal Medicine, Hakuhokai Central Hospital, Hyogo, Japan Abstract. We report that glycated albumin (GA) is higher relative to HbA1c in non-diabetic, gastrecto- mized subjects without anemia, and thus is a sign of oxyhyperglycemia. It is known that gastrectomized subjects are prone to iron-deficiency anemia (IDA), and that the HbA1c levels of subjects with IDAare falsely high. In the present study, the HbA1c and GA levels of gastrectomized subjects with IDA were compared with gastrectomized subjects without anemia. Seven non-diabetic gastrectomized subjects with IDA were enrolled in the present study. Twenty-eight non-diabetic gastrectomized subjects without anemia matched with the subjects with IDA in terms of age, gender, and body mass index were used as the con- trols. Although there were no significant differences in fasting plasma glucose and OGTT 2-hour plasma glucose (2-h PG) between the two groups, the HbA1c and GA levels in gastrectomized subjects with IDA were significantly higher than the controls. For all of the gastrectomized subjects (n=35), ferritin exhibited a significant negative correlation with HbA1c and GA, and a significant positive correlation with2-hPG. In addition, the HbA1c and GA levels exhibited a significant negative correlation with the mean corpuscular hemoglobin and hemoglobin. The HbA1c and GA levels in gastrectomized subjects withIDA were sig- nificantly higher than those in controls. The high GA levels are attributed to a tendency in whichpatients with total gastrectomy, who are prone to IDA, are susceptible to postprandial hyperglycemia and reactive hypoglycemia, which in turn leads to large fluctuations in plasma glucose. Key words: HbA1c, Glycated albumin, Gastrectomy, Iron deficiency anemia, Oxyhyperglycemia. Introduction Glycated albumin (GA) has also been used as an indicator of glycemic control. GA is not affected by The glycation of various proteins is known to occur hemoglobin abnormalities [8], and thus is consid- more often in diabetic patients than non-diabetic ered to be a useful indicator for glycemic control in individuals, and is thought to be involved in the diabetic patients with hemolytic anemia [9] and onset and progression of some chronic diabetic IDA [10]. In addition, it has been shown that GA complications [1]. HbA1c, one of the glycated pro- is a better indicator for plasma glucose fluctuations teins, is commonly used clinically as an indicator of than HbA1c [8]. glycemic control [2,3]. However, HbA1c is also known not to accurately reflect the glycemic con- In gastrectomized subjects, the oral glucose toler- trol in patients with various hematological diseases ance test (OGTT) often shows marked hyperglyce- [4]. The HbA1c levels in patients with hemolytic mia (oxyhyperglycemia) 30 to 60 minutes after glu- anemia characterized by shortened red cell survival cose loading [11,12]. We reported that glycated are low [5]. In contrast, the HbA1c levels in pa- albumin (GA) is higher relative to HbA1c in non- tients with iron-deficiency anemia (IDA) are high diabetic gastrectomized subjects without anemia, [6, 7]. and thus is a sign of oxyhyperglycemia [13]. Gastrectomized subjects are known to complicate Address correspondence to Dr. Masafumi Koga, Department of Internal many types of anemia. One type of anemia that oc- Medicine, Hakuhokai Central Hospital, 4-23-1 Higashisonodacho, curs frequently is IDA, which is induced by a de- Amagasaki, Hyogo 661-0953, Japan; phone: +81 6 4960 6800; fax: +81 6 4960 6900; e mail: [email protected] crease in iron absorption [14,15]. We hypothesized 0091-7370/17/0100-052. © 2017 by the Association of Clinical Scientists, Inc. Labile HbA1c fraction and variant hemoglobin 53 Table 1. Clinical characteristics of the non-diabetic, gastrectomized subjects without anemia and those with iron deficiency anemia (IDA). without anemia with IDA P n 28 7 - Male (%) 19 (68) 4 (57) 0.285 Age (years) 59.0±7.8 59.1±2.2 0.953 BMI (kg/m2) 20.4±1.9 19.5±2.7 0.301 RBC (x104/μl) 444±26 421±28 0.044 Hb (g/dl) 13.9±0.8 11.1±0.9 <0.0001 MCH (pg) 31.2±1.4 26.3±1.3 <0.0001 log ferritin (ng/ml) 1.54±0.36 0.79±0.09 <0.0001 Table 2. COrrelation between indicators of iron deficiency and indicators of glycemic control in the 35 non-diabetic gastrec- tomized subjects. log ferritin (ng/ml) MCH (pg) Hb (g/dl) R P R P R P FPG (mg/dl) 0.067 0.708 –0.007 0.969 0.004 0.983 2-h PG (mg/dl) 0.515 0.002 0.191 0.280 0.212 0.228 HbA1c (%) –0.616 <0.001 –0.503 <0.001 –0.529 0.001 GA (%) –0.320 0.065 –0.437 0.009 –0.604 <0.001 that the HbA1c levels in gastrectomized subjects hemoglobin (MCH) ranges between 28.0 and 34.6 pg for with IDA were higher than those in gastrectomized men, and 26.3 and 34.3 pg for women. This study was subjects without anemia. Therefore, in the present approved by the Ethics Committee of Kinki Central study, HbA1c and GA levels in gastrectomized sub- Hospital, and informed written consent was obtained jects with IDA were compared with those in gastrec- from all patients. tomized subjects without anemia. All subjects had a 75-g OGTT and their glucose tolerance status was diagnosed according to the American Diabetes Materials and Methods Association guidelines [16]. Plasma glucose was deter- mined by the glucose oxidase method. HbA1c, expressed Seven gastrectomized subjects with IDA who exhibited in the National Glycohemoglobin Standardization hypochromic anemia and low ferritin levels (<15 ng/ml) Program (NGSP) value [17], was measured by high per- [four men and three women, 59.1±2.2 years in age, body formance liquid chromatography. GA was determined by mass index (BMI) 19.5±2.7 kg/m2] were enrolled in the the enzymatic method using albumin-specific proteinase, present study. They were selected from among the 99 ketoamine oxidase, and albumin assay reagent (Lucica non-diabetic, gastrectomized subjects, and underwent a GA-L; Asahi Kasei Pharma Co., Tokyo, Japan) [18]. The complete checkup at the Kinki Central Hospital. Twenty- blood cell counts, hematocrit, hemoglobin, and MCH eight non-diabetic, gastrectomized subjects without ane- were measured by using an automated hematology system mia matched for age, gender, and BMI (19 men and 9 (XT-2000i; Sysmex, Kobe, Japan). Serum ferritin concen- women, 59.0±7.8 years in age, BMI 20.4±1.9 kg/m2) trations were measured by using the chemiluminescent were used as the controls. Patients with a history of diabe- immunoassay (CLIA) method (ADVIA Centaur; Siemens tes or evidence of a diabetic pattern on a 75-g oral glucose Medical Solutions Diagnostics, Tokyo, Japan). tolerance test (OGTT) were excluded from the present study. Patients with liver disease and kidney disease, and All the data is shown as mean ± SD. To correct for skewed patients undergoing treatment for anemia were also ex- distributions, serum ferritin levels were logarithmically cluded. The criterion for being diagnosed with anemia transformed. For statistical analysis, paired Student’s t test was an Hb level of less than 13 g/dl for men, and less than or chi-square test was used to compare two groups when 12 g/dl for women. The criterion for mean corpuscular deemed appropriate. To analyze the effects of explanatory 54 Annals of Clinical & Laboratory Science, vol. 47, no. 1, 2017 Figure 1. COmparison between fasting plasma glucose (FPG) and OGTT 2-hour plasma glucose (2-h PG) in the twO groups. The figure shows FPG (A) and 2-h PG (B) in the gastrecto- mized subjects without anemia (open columns) and in the gastrecto- mized subjects with IDA (closed columns). variables on log-transformed ferritin, MCH, and Hb, was also observed, but this was not significant. In univariate regression analysis was performed using the addition, Hb and MCH exhibited a significant StatView computer program (Version 5.0 for Windows, negative correlation with HbA1c and GA (Table Abacus Concepts, Berkeley, CA). A P-value of <0.05 was 2). Neither Hb nor MCH exhibited a significant considered to be statistically significant. correlation with FPG or 2-h PG. Results Discussion The red blood cell counts in gastrectomized sub- The indicators of glycemic control in non-diabetic, jects with IDA were slightly less than those in the gastrectomized subjects with IDA were compared controls. The Hb, MCH, and ferritin levels were with those of the controls, who were the non-dia- significantly welo r than those in the controls (Table betic, gastrectomized subjects without anemia. The 1). No significant difference was observed between HbA1c levels in gastrectomized subjects with IDA the two groups in fasting plasma glucose (FPG). were significantly higher than those in the controls. The OGTT 2-hour plasma glucose (2-h PG) levels For all gastrectomized subjects, HbA1c exhibited a of gastrectomized subjects with IDA exhibited no significant negative correlation with ferritin, MCH, significant difference with those in the controls, but and Hb, the indicators of iron deficiency. These re- tended to be low (Figure 1). The HbA1c levels in sults suggested falsely high HbA1c levels due to gastrectomized subjects with IDA were significantly IDA in gastrectomized subjects. Although GA is higher than the controls (6.2±0.4% vs. 5.7±0.3%; known to be not influenced by anemia, the GA lev- P=0.003) (Figure 2).