CLINICAL INVESTIGATION

Ischemia Modified Albumin Test to

Detect Early Diabetic Complications
Kinjalka Ghosh, MD, Manohar G. Muddeshwar, PhD and
Kanjaksha Ghosh, MD

ABSTRACT
Background: The objective of this study was to assess the albumin cobalt binding (ACB) test in a cohort of type 2 diabetes
patients. The ACB test is a simple, inexpensive, sensitive and robust test that could have important clinical application in
detecting complications of type 2 diabetes mellitus.
Materials and Methods: We tested patients with type 2 diabetes without any clinically detectable complications or without
any other comorbid conditions for serum ACB levels along with an equal number of age- and sex-matched healthy control
subjects. ACB levels were compared after the patients with diabetes were investigated for various complications using
standard statistical tests of significance.
Results: A total of 100 patients with type 2 diabetes were studied with age- and sex-matched healthy control subjects.
Of the 100 patients, 78 had different complications on detailed laboratory testing. The patients with complications had
significantly higher ACB test results when compared to the patients with diabetes without complications and to that of the
control subjects (0.62 ⫾ 0.04, 0.42 ⫾ 0.07, 0.30 ⫾ 0.05 absorbance units (ABSU)/mL, respectively. P o 0.001). All values in
diabetics were significantly higher than that of controls.
Conclusions: The serum ACB test is a sensitive indicator of complications developed in type 2 diabetes mellitus. Patients
may be followed up with ACB results to detect early complications in this disease.
Key Indexing Terms: Albumin-cobalt-binding assay; Ischemia modified albumin; Diabetes; Complications; Ultrasensitive
C-Reactive protein. [Am J Med Sci 2017;354(5):467–470.]

INTRODUCTION disease, unstable angina and acute myocardial infarction

(MI). This test is also known as ischemia modified

C
ertain investigations in medical sciences are non-
specific, but provide the clinician with further
impetus to explore the disease conditions in
which the results of those investigations are abnormal.
Even normalization of the results in those investigations
has a prognostic value. The management strategies can
be adopted to treat those conditions using these kind of
tests to ensure that the treatment is progressing well.
Erythrocyte sedimentation rate, serum C-reactive protein
(CRP), serum procalcitonin and serum lactate dehydro-
genase levels are examples of such investigations.
Serum albumin is the most abundant plasma protein
and binds many compounds including organic and
inorganic cations and anions. Some of these bindings
also have allosteric effects on binding of other important
ligands.1
It has been demonstrated that albumin binds tran-
sition metal ions like nickel, cobalt and copper with
some affinity, and this binding affinity is lost when the
albumin molecule is damaged in a certain way2,3 due to
various disease conditions. Originally the albumin-cobalt
binding (ACB) technique was developed in the form of a
simple spectrophotometric assay,4 and was found to
detect ischemic heart disease and acute myocardial
information 6-24 hours before troponin T levels in the
serum began to rise.5 It reliably discriminated between
nonspecific chest pain from that of ischemic heart
albumin assay (IMA) and is available in several formats
(simple spectrophotometric, immunochemical and com-
plex assays). The test has also been evaluated as a
sensitive biomarker in several conditions.6,7
We evaluated the value of the albumin cobalt binding
test for IMA levels in serum of patients with type 2
diabetes mellitus for detection of early diabetic
complications.
MATERIAL AND METHODS
A total of 100 consecutive patients with type 2
diabetes were included in the study. Patients had been
diagnosed with type 2 diabetes for 6 months to 5 years
and had no apparent complications as evidenced by
clinical examination. The same number of age- and
sex-matched healthy volunteers were included for
comparison.
The study was sanctioned by the Institutional Review
Board (Indira Gandhi Government Medical College,
Nagpur) and both the patients and the controls provided
informed consent.
The data for detailed clinical proforma were gathered
for all patients and controls. Investigations included
fasting (12 hours) blood sugar, lipid profile, HbA1c levels,
ultrasensitive CRP levels, 12-lead electrocardiogram

Copyright © 2017 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. 467
www.amjmedsci.com
www.ssciweb.org
Ghosh et al

with rhythm strip, liver function test, renal function test, TABLE 1. Parameters in patients and controls.
routine urine examination, albumin/creatine ratio in the
Patients Controls
urine and abdominal ultrasonography for all the patients Parameters (n ¼ 100) (n ¼ 100)
and controls. In addition, both the patients and controls
had a 2-hour postprandial blood glucose estimation after Age (years) 39-74 38-74
75 g of oral glucose following overnight fasting. The (mean ⫾ 1 SD) (56.2 ⫾ 9.4) (57.6 ⫾ 9.4)
patients and controls also had further examination by an Male:female 57:43 57:43
ophthalmologist after dilating the pupil with 0.5% tropi- BMI 26.1 ⫾ 2.6 22.8 ⫾ 1.5
camide. Diabetic patients additionally underwent fluo- Waist circumference (cm) 91 ⫾ 9 81 ⫾ 9
Serum Albumin (g/dL) 4.1 ⫾ 0.56 4.0 ⫾ 0.46
rescein angiography and nerve conduction velocity
Liver function tests Normal Normal
studies.
Serum creatinine (mg/dL) 1.03 ⫾ 0.24 0.70 ⫾ 0.22
The ACB test for IMA was conducted as per Bar-Or.4
Fasting blood sugar (mg/dL) 134 ⫾ 9.5 87.8 ⫾ 11
We added 200 mL of serum to 50 mL of CoCl2 (1 g/L)
2-hour postprandial (mg/dL) 228.6 ⫾ 39 121 ⫾ 9
followed by vigorous mixing and 10 minutes incubation. (75 g glucose load)
Dithiothreitol (1.5 g/L) 50 mL was then added to the HbA1c level (%) 7.8 ⫾ 1.6 6.2 ⫾ 0.05
mixture for development of color. After 2 minutes, the Serum triglyceride (mg/dL) 164 ⫾ 30 148 ⫾ 29
reaction was terminated by adding 1 mL of 9 g/L of HDL (mg/dL) 33 ⫾ 7 39 ⫾ 7
NaCl. The absorbance of the assay mixture was taken at LDL (mg/dL) 150 ⫾ 33 80 ⫾ 22
470 nm in a spectrophotometer. All reagents were VLDL (mg/dL) 33 ⫾ 7 30 ⫾ 6
purchased from Sigma Chemicals (St. Louis, MO). Albumin in urine 113 ⫾ 109 16 ⫾ 4
Standard statistical tests of , that is, chi-square (χ2) (mg/g of creatinine)
test, Student’s t test and one-way ANOVA test were Complete hemogram Normal Normal
carried out to establish the statistical significance of the High sensitivity CRP (mg/L) 3.1 ⫾ 1.3 0.71 ⫾ 0.3
findings. Any value with P o 0.05 was considered
All values are in mean ⫾ 1 SD. Except age, sex, serum albumin levels and
significant. A z-test and receiver operating characteristic hemogramall other values show a P o 0.001 on t test between cases and
curve analysis were conducted to develop the cutoff controls. BMI, body mass index; CRP, C-reactive protein; HDL, high-
values of the ACB test for IMA, expressed as absorbance density lipoprotein; LDL, low-density lipoprotein; VLDL, very low-density
units (ABSU)/mL, and also to determine the sensitivity lipoprotein.

and specificity of the test in detecting diabetic

complications.
RESULTS
Patients were between 39 and 74 years with a
mean ⫾ 1 SD of 57.6 ⫾ 9.4 years. It was comparable
to that of the controls, which were between 38 and 74
years with a mean ⫾ 1 SD of 56.2 ⫾ 9.4 years. Male-to-
female ratio in both the groups was the same (57:43). The
patients had significantly higher fasting and postprandial
blood sugar levels (134 ⫾ 9.5 mg/dL and 228 ⫾ 39 mg/
dL). The HbA1c levels of the patients were 7.8 ⫾ 1.6%.
Higher fasting serum triglyceride, low-density lipoprotein
and very low-density lipoprotein levels and lower high-
density lipoprotein levels (P o 0.001) were observed
among the patients. The body mass index and waist
circumference in the patients and controls were also
significantly different (26.1 ⫾ 2.6 versus 22.8 ⫾ 1.5 for
body mass index and 91 ⫾ 9 cm versus 81 ⫾ 8 cm for
waist circumstance; P o 0.001 for both).
Ultrasensitive CRP assay (Tulip diagnostics, Goa,
India) was also significantly higher in the patients
compared to control (3.1 ⫾ 1.3 mg/L versus 0.71 ⫾
0.03 mg/L, P o 0.001). Urine albumin and creatinine
levels as well as serum creatinine levels were also higher
in patients with diabetes (113 ⫾ 108 mg versus 16 ⫾
4 mg of albumin/g of creatinine in urine and serum
creatinine of 1.03 ⫾ 0.24 mg/dL versus 0.70 ⫾
0.22 mg/dL for both analysis; P o 0.001).
None of the control subjects had microalbuminuria,
whereas 42 patients with diabetes had microalbuminuria
(30-300 mg albumin/g of creatinine) and 5 patients had
macroalbuminuria (4300 mg albumin/g of creatinine). All
these investigations have been summarized in Table 1.
Ophthalmoscopic examinations of the controls were
essentially normal.
The ACB test was performed with plasma from 30
healthy volunteers (different from those controls of this
study) to develop a reference range for this test. Multiple
assays by different technicians were performed with a
mean of 0.30 ABSU/mL with intra-assay and inter-assay
coefficient variation of 3.7% and 6.4%. All the patients
and controls had normal hemoglobin, normal liver func-
tion and renal functional tests. Of the 100 patients with
diabetes in our study, 78 were found to have complica-
tions of diabetes either singly or in combination (pro-
liferative retinopathy, neuropathy or nephropathy),
presented in Table 2. We have not included the 30
volunteers whose blood was used to develop the
reference range of ACB value in our laboratory for
comparison analysis with diabetic patients, as these
healthy volunteers did not have all the investigations
done, which were done for the controls of the present
study. However, even if these values were merged with
the 100 controls used in this study the mean and
standard deviation for ACB in 130 controls showed
hardly any change.

468 THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

VOLUME 354 NUMBER 5 November 2017
 Ischemia Modified Albumin in Early Diabetic Complications

TABLE 2. Levels of ACB and ultrasensitive CRP in diabetic complications. of the results in diabetic retinopathy showed it to be a
test of great sensitivity to pick up proliferative retinop-
Complications ACB (ABSU/mL, US CRP (mg/L,
(n) mean ⫾ 1 SD)* mean ⫾ 1 SD)
athy.14 However, a study by D’Souza et al15 did not find
good correlation with IMA (ACB) levels and diabetic
Retinopathy (17) 0.59 ⫾ 0.04 3.00 ⫾ 0.61 complications, though he found a correlation with
Neuropathy (8) 0.62 ⫾ 0.04 3.06 ⫾ 1.19 HbA1c levels. There could be several reasons for these
Nephropathy (15) 0.61 ⫾ 0.04 2.70 ⫾ 0.46 discrepant findings compared to our study. The patients
Retinopathy and 0.64 ⫾ 0.04 4.50 ⫾ 1.15 in D’Souza’s group probably had multiple comorbidities,
neuropathy (6)
which may have diluted the effect of IMA (ACB) levels in
Nephropathy and 0.64 ⫾ 0.05 3.67 ⫾ 1.04
neuropathy (7)
detecting the diabetic complications. In the present
Retinopathy and 0.63 ⫾ 0.02 3.94 ⫾ 0.75
study, we evaluated the ACB test in parallel with the
nephropathy (13) ultrasensitive CRP test, for picking up early diabetic
All complications (12) 0.66 ⫾ 0.03 4.79 ⫾ 1.12 complications. We excluded patients with clear cut
Total with 0.62 ⫾ 0.04 3.61 ⫾ 1.1 clinically evident diabetic complications such as asso-
complications (78) ciated clinically significant atherosclerotic disorders in
No complication (22) 0.42 ⫾ 0.07 1.6 ⫾ 0.4 the form of abnormal ECGs, previous strokes, myocar-
Controls (100) 0.30 ⫾ 0.05 0.75 ⫾ 0.03 dial infarctions, clinically abnormal pulses, diabetic foot,
P o 0.001 P o 0.001 gangrene and other comorbidities.
ANOVA test, F ¼ 368.959, degree of freedom 271. P o 0.001. Patients on various combinations of lipid lowering drugs
ABSU=mL could also be calculated as : *ABSU=mL as above table
serum albumin in g=dL  1000
were also excluded. With that kind of patient population the
This corrects the value for extremes of serum albumin levels which ACB test was found to clearly detect the early diabetic
was not required in our patient series. ABSU, absorbance units; ACB,
complications with high sensitivity and specificity and is
albumin-cobalt binding; SD, standard deviation; US CRP, ultrasensitive C-
reactive protein. almost similar to the ultrasensitive CRP assay.
Ultrasensitive CRP assay is a costly investigation
compared to the ACB assay and is often not available
Serum ACB levels in patients and controls were 0.58 outside routine hours in Indian hospitals. ACB test can
⫾ 0.09 ABSU/mL versus 0.30 ⫾ 0.05 ABSU/mL (P o be carried out at a low cost, as a single test or as a batch
0.001). When the patients with diabetes were assessed processing test, and in addition to diabetes may be
for the ACB test, it was observed that patients with evaluated in many other pathological conditions. It
complications had significantly elevated levels of IMA remains to be seen if in real world situations where
than those without them (0.62 ⫾ 0.04 ABSU/mL versus significant comorbidities and multiple drug therapies are
0.42 ⫾ 0.06 ABSU/mL, P o 0.001). often associated with patients with type 2 diabetes this
Receiver operator characteristic curve for ACB in test will perform well or not. Associated comorbidities
detecting complications of diabetes gave an AUC of are likely to reduce the predictive values of the test,
0.993 ⫾ 0.005 with an optimum cutoff level of particularly the positive predictive value. However, a
0.54 ABSU/mL at a specificity of 97.44% and sensitivity normal ACB test is likely to be associated with the
of 95.45%. Taking 0.54 ABSU/mL as a cutoff level absence of diabetic complications. Hence it may still be
resulted in correct detection of 76 of the 78 patients useful as a simple test for follow-up investigation in
who had complications (χ2 ¼ 83.61, P o 0.001). diabetic patients as an early test for oncoming compli-
With ultrasensitive CRP levels at a cutoff value of cations, then proper investigations and management
2.1 mg/dL, the sensitivity and specificity to pick up strategies can follow.
diabetic complications were found to be 94.387% and
95.4%, respectively.
REFERENCES
1. Blindauer CA, Khazaipoul S, Yu R, et al. Fatty acid-mediated inhibition
DISCUSSION of metal binding to the multi-metal site on serumalbumin: implications for
cardiovascular disease. Curr Top Med Chem 2016;16:3021–32.
ACB or IMA have been evaluated as a prognostic 2. Bar-Or D, Rael LT, Bar-Or R, et al. The cobalt-albumin binding assay:
marker in several disease conditions6,7 where the oxidant insights into its mode of action. Clin Chim Acta 2008;387(1-2):120–7.
or inflammatory damage to the tissue modify circulating 3. Sbarouni E, Georgiadou P, Voudris V. Ischemia modified albumin
albumin molecules and alters its affinity for transition changes—review and clinical implications. Clin Chem Lab Med 2011;49:
177–84.
elements like cobalt, nickel and copper. The test was
4. Bar-Or D, Lau E, Winkler JV. A novel assay for cobalt albumin binding
found to detect acute myocardial infarction well before and its potential as a marker for myocardial ischemia—a preliminary
the rise in troponin T levels in patients who presented report. J Emerg Med 2000;19:311–5.
with chest pain in the emergency department.5,8,9 5. Christenson RH, Duh SH, Sanhai WR, et al. Characteristics of an
The test is inexpensive (o20 cent/test), robust and albumin cobalt binding test for assessment of acute coronary syndrome
patients: a multicenter study. Clin Chem 2001;47:464–70.
easy to perform and can be done in 15 minutes.10 This 6. Chen CY, Tsai WL, Lin PJ, et al. The value of serum ischemia-modified
test has been evaluated by several investigators for albumin for assessing liver function in patients with chronic liver disease.
various diabetic complications11-13 and a meta-analysis Clin Chem Lab Med 2011;49:1817–21.

Copyright © 2017 Southern Society for Clinical Investigation. Published by Elsevier Inc. All rights reserved. 469
www.amjmedsci.com
www.ssciweb.org
Ghosh et al
7. Ashok Kumar P, Anand U. Multiple biomarkers to assess the patho-
physiological state in critically ill patients with sepsis. Indian J Clin Biochem
2016;31:310–4.
8. Reddy CB, Cyriac C, Desle HB. Role of ischemia modified albumin (ima)
in acute coronary syndromes. Ind Heart J 2014;66:656–62.
9. Abadie JM, Blassingame CL, Bankson DD. Albumin cobalt binding assay
to rule out acute coronary syndrome. Ann Clin Lab Sci 2005;35:66–72.
10. Chawla R, Goyal N, Calton R, et al. Ischemia modified albumin: a novel
marker for acute coronary syndrome. Ind J Clin Biochem 2006;21:77–82.
11. Ma SG, Jin Y, Xu W, et al. Increased serum levels of ischemia-modified
albumin and C-reactive protein in type 1 diabetes patients with ketoaci-
dosis. Endocrine 2012;42:570–6.
12. Kaefer M, Piva SJ, De Carvalho JA, et al. Association between
ischemia modified albumin, inflammation and hyperglycemia in type 2
diabetes mellitus. Clin Biochem 2010;43:450–4.
13. Ahmad A, Manjrekar P, Yadav C, et al. A.evaluation of ischemia-
modified albumin, malondialdehyde, and advanced oxidative protein
products as markers of vascular injury in diabetic nephropathy. Biomark
Insights 2016;11:63–8.
14. Reddy VS, Sethi S, Gupta N, et al. Significance of ischemia-modified
albumin as a simple measure of oxidative stress and its discriminatory
470
ability in diabetic retinopathy : literature review and meta analysis. Retina
2016;36:1049–57.
15. D’Souza JM, D’Souza RP, Vijin VF, et al. High predictive ability of
glycated hemoglobin on comparison with oxidative stress markers in
assessment of chronic vascular complications in type 2 diabetes mellitus.
Scand J Clin Lab Invest 2016;76:51–7.
From the Department of Clinical Biochemistry (KG, MGM), Indira Gandhi
Government Medical College Hospital, Maharashtra, India; Surat Raktadan
Kendra & Research Centre (KaG), Gujrat, India.
Submitted March 28, 2017; accepted June 21, 2017.
The authors have no financial or other conflicts of interest to disclose.
All costs were borne by the investigators.
Present address (KG): Department of Clinical Biochemistry, Seth GS
Medical College & KEM hospital, Parel, Mumbai 400012, Maharashtra, India.
Correspondence: Kanjaksha Ghosh, MD, DNB, FRCPath, FACP, Surat
Raktadan Kendra & Research Centre, Udhna Magdalla Road, Near Chosath
Joganio Mata Mandir, Surat 395002, Gujrat, India (E-mail: kanjaksha-
ghosh@hotmail.com).
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
VOLUME 354 NUMBER 5 November 2017