CLB-09021; No. of pages: 4; 4C: Clinical Biochemistry xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Short Communication
GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose tolerance status Robin P.F. Dullaart a,⁎, Eke G. Gruppen a,b, Margery A. Connelly c, James D. Otvos c, Joop D. Lefrandt d a
Department of Endocrinology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands Department of Nephrology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands LabCorp, Raleigh, NC, USA d Department of Vascular Medicine, University of Groningen, University Medical Center Groningen, The Netherlands b c
a r t i c l e
i n f o
Article history: Received 27 March 2015 Received in revised form 1 May 2015 Accepted 2 May 2015 Available online xxxx Keywords: C-reactive protein Glycoproteins Diabetes mellitus Leptin/adiponectin ratio Nuclear magnetic resonance spectroscopy
a b s t r a c t Objectives: Plasma GlycA is a recently developed biomarker whose nuclear magnetic resonance signal originates from glycosylated acute-phase proteins. The aim of our study was to determine potential relationships between GlycA and adiposity, insulin resistance (HOMAir), high sensitive C-reactive protein (hs-CRP), leptin, adiponectin, and the leptin/adiponectin ratio, and to test whether GlycA is elevated in subjects with impaired fasting glucose (IFG) and type 2 diabetes mellitus (T2DM). Design and methods: Plasma GlycA, hs-CRP, leptin, adiponectin, the leptin/adiponectin ratio, and insulin resistance (HOMAir) were measured in 103 fasting subjects (30 with normal fasting glucose, 25 with IFG and 48 with T2DM). Results: In all subjects combined, plasma GlycA was correlated positively with body mass index (BMI), HOMAir, hs-CRP, leptin and the leptin/adiponectin ratio, and inversely with adiponectin (p b 0.05 to p b 0.001). GlycA did not significantly vary according to glucose tolerance category (p = 0.060). GlycA was related positively to the leptin/adiponectin ratio (p = 0.049), independent of BMI (p = 0.056) and HOMAir (p = 0.50). Conclusions: High plasma GlycA reflects a pro-inflammatory state. Adipose tissue-associated inflammatory processes could contribute to increased circulating levels of glycosylated acute-phase proteins. © 2015 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Protein glycosylation, i.e. the enzymatic process whereby a glycan (carbohydrate) moiety is linked to a protein, is influenced by many biological processes including inflammation [1]. Proton nuclear magnetic resonance (NMR) spectroscopy has the ability to detect signals from circulating glycoproteins [2]. Interestingly, the NMRbased glycoprotein biomarker, designated GlycA, which captures both the protein levels and enhanced glycosylation states of the most abundantly expressed acute-phase proteins, has been shown to predict the development of cardiovascular disease (CVD), as well as the progression to type 2 diabetes mellitus (T2DM) in women [3,4]. Inflammatory processes are characterized by enhanced acute phase protein secretion and glycosylation [5]. Accordingly, a positive relationship of GlycA with high-sensitive C-reactive protein (hs-CRP) has been observed [2,3], but insights into the relationships between GlycA and other cardiometabolic risk factors are still limited. Among adipokines, leptin and adiponectin play key roles in adipocyte differentiation and ⁎ Corresponding author at: Department of Endocrinology, University Medical Center Groningen, University of Groningen, P.O. Box 30.001, Groningen 9700 RB, The Netherlands. Fax: +31 503619392. E-mail address: [email protected] (R.P.F. Dullaart).
inflammation [6,7]. The leptin/adiponectin ratio is a determinant of insulin resistance in non-diabetic individuals, highlighting the contribution of adipose tissue dysfunction to the pathogenesis of diminished insulin action [8]. Furthermore, a high leptin/adiponectin ratio may confer increased intima media thickness and predict incident CVD [9–11]. Given that GlycA may predict CVD and progression to T2DM, we hypothesized that GlycA levels may be related to disturbances in pro- and anti-inflammatory adipokines. Therefore, the aim of this study was to interrogate potential relationships of GlycA between and hs-CRP, adiposity, insulin resistance, leptin, adiponectin and the leptin/adiponectin ratio. Second, we tested whether plasma GlycA is elevated in subjects with varying degrees of glucose intolerance. Materials and methods Subjects The medical ethics committee of the University Medical Center Groningen approved the study protocol. Participants with and without T2DM were aged N18 years and were recruited by advertisement in local newspapers. Written informed consent was obtained. T2DM had been diagnosed previously (fasting plasma glucose ≥ 7.0 mmol/L; non-fasting plasma glucose ≥11.1 mmol/L). In non-diabetic subjects,
http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001 0009-9120/© 2015 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
2
R.P.F. Dullaart et al. / Clinical Biochemistry xxx (2015) xxx–xxx
glucose tolerance status was classified as normal fasting glucose (NFG; plasma glucose b5.6 mmol/L) or impaired fasting glucose (IFG; plasma glucose ≥ 5.6 and ≤ 6.9 mmol/L), according to NCEP-ATPIII criteria. T2DM subjects using insulin, current smokers and subjects who used lipid lowering drugs were excluded, as were subjects with a history of CVD, chronic kidney disease, liver function abnormalities or thyroid dysfunction. Subjects using metformin, sulfonylurea or antihypertensive medication were allowed to participate. Blood pressure was measured after 15 min rest at the left arm in sitting position using a sphygmomanometer. Mean arterial pressure was calculated as 1 / 3 × systolic blood pressure + 2 / 3 × diastolic blood pressure. Body mass index (BMI in kg/m2) was calculated as weight divided by height squared. Homeostasis model assessment of insulin resistance (HOMAir) was used to estimate insulin resistance (fasting insulin (mU/L) × fasting glucose (mmol/L) / 22.5).
and the leptin/adiponectin ratio were used. Differences in continuous variables according to glucose tolerance status category were assessed by one-way analysis of variance (ANOVA) with Bonferroni correction for multiple comparisons. Differences in proportions in categorical variables between the glucose tolerance groups were assessed by χ2analysis. The relationship of GlycA and hs-CRP with T2DM was also determined using receiver operating characteristic (ROC) analysis, presented as the area under the curve (AUC) with 95% confidence intervals (CI). Univariate relationships were assessed with Pearson correlation coefficients. Partial correlation coefficients were calculated after adjustment for glucose tolerance category, age and sex. Multivariable linear regression analyses were carried out to disclose the independent contribution of variables to GlycA and hs-CRP. Two-sided p-values b0.05 were considered significant. Results
Laboratory methods Venous blood was collected in EDTA containing tubes. Plasma was stored at −80 °C until analysis. Plasma glucose was measured shortly after blood collection with an APEC glucose analyzer. Total cholesterol and triglycerides were measured by routine enzymatic methods. HDL cholesterol was assayed by a homogeneous enzymatic colorimetric test. NMR spectra were obtained from plasma samples using the NMR Profiler [12]. NMR signal amplitudes originating from the N-acetyl methyl group protons of the N-acetylglucosamine moieties located on the bi-, tri-, and tetra-antennary branches of plasma proteins, predominantly α1-acid glycoprotein, haptoglobin, α1-antitrypsin, α1antichymotrypsin, and transferrin, were used to calculate the GlycA concentration (μmol/L of N-acetyl methyl groups) [2,3]. The intra- and inter-assay coefficients of variation (CV) were 1.9% and 2.6%, respectively [2]. Leptin and total adiponectin were assayed as described [13]. Intra-assay and inter-assay CVs were b 6% and b8%, respectively. Insulin, hs-CRP and glycated hemoglobin (HbA1c) were determined as described [13]. Statistical analysis SPSS (version 20.0, SPSS Inc. Chicago, IL, USA) was used for data analysis. Results are expressed as mean ± SD or as median (interquartile range). Because of skewed distributions, logarithmically transformed values of triglycerides, insulin, HOMAir, hs-CRP, leptin, adiponectin,
One hundred and three subjects participated, of whom 30 were classified with NFG and 25 with IFG (Table 1). Of the 48 subjects with T2DM, 11 used metformin, 9 used sulfonylurea and 15 used both drugs. Antihypertensive medication (mostly angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, diuretics and βblockers, alone or in combination) were used by 23 T2DM subjects. Glucose lowering drugs and antihypertensive medication were not used by non-diabetic subjects. One postmenopausal woman with T2DM, 1 postmenopausal woman with IFG 1, and 1 premenopausal woman with NFD used estrogens. Other medications were not taken. T2DM subjects were older, had higher blood presssure, were more obese and more insulin resistant compared to NFG and IFG subjects (Table 1). There were no gender differences between the groups. Total cholesterol did not differ between the groups, but triglycerides were higher and HDL cholesterol was lower in T2DM than in IFG and NFG subjects. Leptin and the leptin/adiponectin ratio were higher, whereas adiponectin was lower in T2DM compared to IFG and NFG subjects. hs-CRP was higher in T2DM than in IFG subjects. The difference in GlycA between the 3 glucose tolerance groups was not significant. There was also no difference in GlycA between the glucose tolerance groups after age and sex adjustment (p N 0.30). However, the AUC of the ROC curve for GlycA with T2DM vs. no T2DM as the outcome was significant (0.624 (95% CI), 0.515–0.733; p = 0.030). For comparison, the AUC for hs-CRP with T2DM was 0.662 (95% CI, 0.558–0.766; p = 0.005). For GlycA, equal sensitivity and specificity (0.574) were reached
Table 1 Clinical characteristics, plasma glucose, insulin, insulin resistance, plasma lipoproteins, adipokines, high sensitivity C-reactive protein and GlycA in subjects with normal fasting glucose (n = 30), impaired fasting glucose (n = 25) and type 2 diabetes mellitus (n = 48).
Age (years) Sex (men/women) MAP (mm Hg) BMI (kg/m2) Waist circumference (cm) Fasting glucose (mmol/L) HbA1c (mmol/mol) Fasting insulin (mU/L) HOMAir (mU × mmol/L2 × 22.5) Total cholesterol (mmol/L) HDL cholesterol (mmol/L) Triglycerides (mmol/L) Leptin (μg/L) Adiponectin (mg/L) Leptin/adiponectin ratio (μg/mg) GlycA (μmol/L) hs-CRP (mg/L)
Normal fasting glucose (n = 30)
Impaired fasting glucose (n = 25)
Type 2 diabetes mellitus (n = 48)
p-Value*
52 ± 10b 11/19 99.3 ± 15.2b 25.9 ± 4.6c 85.2 ± 13.0c 5.2 ± 0.3c 34 ± 3c 6 (4–9)c 1.4 (1.0–2.0)c 5.7 ± 1.1 1.5 ± 0.4b 1.33 (0.85–2.03)b 6.8 (3.2–22.7)a 22.7 (15.0–48.1)b 0.34 (0.11–0.74)c 373 ± 48 1.81 (0.59–2.92)
56 ± 9 11/14 96.9 ± 10.9c 25.0 ± 3.0c 87.2 ± 10.5c 6.1 ± 0.4c 33 ± 0.3c 6 (5–8)c 1.7 (1.2–2.3)c 5.8 ± 0.8 1.6 ± 0.4c 1.34 (0.92–1.88)b 5.9 (3.3–23.7)a 21.3 (16.2–40.7)b 0.31 (0.13–0.67)c 364 ± 60 0.85 (0.47–1.90)b
59 ± 9 28/20 108.7 ± 10.4 30.7 ± 4.5 106.7 ± 12.6 9.4 ± 2.3 51 ± 8 12 (9–18) 5.7 (3.5–7.7) 5.5 ± 0.9 1.2 ± 0.3 1.94 (1.49–2.54) 14.7 (6.8–31.9) 14.4 (9.5–26.7) 0.97 (0.39–2.13) 395 ± 52 2.19 (1.36–4.20)
0.003 0.15 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.22 b0.001 0.001 0.01 b0.001 b0.001 0.06 0.002
Data in mean ± SD or in median (interquartile range). BMI: body mass index; HbA1c: glycated hemoglobin; HDL: high density lipoproteins; HOMAir: homeostasis model assessment of insulin resistance; hs-CRP: high sensitive C-reactive protein; MAP: mean arterial pressure. ap b 0.05 vs. subjects with type 2 diabetes mellitus; p-value*: p-value by analysis of variance; bp b 0.01 vs. subjects with type 2 diabetes mellitus; cp b 0.001 vs. subjects with type 2 diabetes mellitus.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
R.P.F. Dullaart et al. / Clinical Biochemistry xxx (2015) xxx–xxx
at a value of 383 μg/L, whereas for hs-CRP equal sensitivity and specificity (0.553) were reached at a value of 1.75 mg/L. In all subjects combined, GlycA was correlated positively with BMI, insulin, hs-CRP, HOMAir, leptin and the leptin/adiponectin ratio, and inversely with adiponectin (Supplementary Table 1A). hs-CRP was also correlated positively with BMI, insulin, HOMAir, leptin and the leptin/ adiponectin ratio, and inversely with adiponectin (Supplementary Table 1A). Similar relationships were observed in partial correlation analysis taking account of age, sex, and glucose tolerance category (Supplementary Table 1B). GlycA was correlated positively with hsCRP in the 3 glucose tolerance groups separately (NFG: r = 0.427, p b 0.01; IFG: r = 0.741, p b 0.001; T2DM: r = 0.555, p b 0.001). Neither GlycA, nor hs-CRP were correlated with plasma glucose (Supplementary Tables 1A and B). Multivariable linear regression analysis was performed to determine the extent to which GlycA and hs-CRP were associated with leptin, adiponectin, and the leptin/adiponectin ratio when taking BMI and HOMAir into account (Table 2). In age- and sex-adjusted analyses, GlycA was related to BMI but not significantly to HOMAir (Table 2A Model 1). GlycA was not independently associated with leptin and adiponectin as separate variables (Model 2). Of note, GlycA was positively related to the leptin/adiponectin ratio, independently of BMI and HOMAir (Model 3). hs-CRP was associated with BMI (Table 2B, Model 1), as well as with leptin (Model 2) and the leptin/ adiponectin ratio (Model 3). The strength of the association of GlycA with the leptin/adiponectin ratio was comparable to that of hs-CRP with the leptin/adiponectin ratio (Table 2A Model 3 vs. Table 2B Model 3). When glucose tolerance status, the use of metformin, sulfonylurea and antihypertensive medication were additionally taken into account, the relationship of GlycA with the leptin/adiponectin ratio (cf. Table 2A, Model 3; β = 0.290, p = 0.032), as well as of hsCRP with the leptin/adiponectin ratio (cf. Table 2B, Model 3; β = 0.293, p = 0.019) remained essentially unaltered.
Discussion This study reveals a strong positive relationship between an NMRbased biomarker of inflammatory glycoproteins, designated GlycA, and hs-CRP. Like hs-CRP, GlycA was independently associated with the leptin/adiponectin ratio, a finding which reinforces recent observations
3
demonstrating that this glycoprotein biomarker signal reflects a proinflammatory state [2–4]. Metabolic profiling holds promise for delineating hitherto unappreciated pathways which may conceivably contribute to the development of atherosclerosis [3,14]. Specific glycan structures of acute-phase proteins are enzymatically modified under chronic inflammatory conditions [1,5]. This suggests that GlycA, which represents an integrated plasma glycoprotein biomarker, reflects a pro-inflammatory state beyond associations with the individually captured proteins. In line with this supposition, an NMR-based risk score, including one of the proteins captured in the GlycA NMR signal, namely α-1-acid glycoprotein, may predict all cause and CVD mortality [14]. Well-documented effects of leptin and adiponectin on inflammatory processes [6,7] prompted us to determine the associations of these adipokines with GlycA in comparison with hs-CRP. As shown in the current report, GlycA was related positively to the leptin/ adiponectin ratio, extending expected associations of hs-CRP with this adipokine ratio [10,11]. Of potential relevance, the relationships of GlycA and hs-CRP with the leptin/adiponectin ratio were independent from BMI, a novel finding which agrees with the hypothesis that abnormalities in adipose tissue function could contribute to a proinflammatory state [8]. Several methodological aspects of our study need further discussion. The cross-sectional design of our study does not allow to address the nature of the observed relationships. For this reason, we cannot exclude the possibility of reverse causation. Furthermore, GlycA did not significantly vary according to the degree of glucose intolerance, although ROC analysis indicated that a significant relation of GlycA with the presence of T2DM. In addition, there was no independent association of GlycA with HOMAir, a validated estimate of insulin resistance also in T2DM [15]. In the interpretation of these results it should be considered that glycemic control was adequate in most T2DM subjects. Therefore, the possible effect of more severe hyperglycemia in elevating GlycA warrants further evaluation. Finally, it should be recognized that the enzymatic process of protein glycosylation has to be discerned from protein glycation, a non-enzymatic process which is accelerated under hyperglycemic circumstances. In conclusion, the glycan biomarker, GlycA, is associated positively with the leptin/adiponectin ratio. It is conceivable that adipose tissueassociated inflammatory processes could contribute to increased levels of glycosylated acute-phase proteins.
Table 2 Multivariable linear regression analyses demonstrating independent relationships of GlycA (A) and high sensitive C-reactive protein (hs-CRP) (B) with body mass index (BMI), homeostasis model assessment of insulin resistance (HOMAir), leptin, adiponectin, and the leptin/adiponectin ratio. Model 1 β A GlycA Age Sex (men/women) BMI HOMAir Leptin Adiponectin Leptin/adiponectin ratio B hs-CRP Age Sex (men/women) BMI HOMAir Leptin Adiponectin Leptin/adiponectin ratio
Model 2 p-Value
β
Model 3 p-Value
β
p-Value
0.204 −0.018
0.038 0.86
0.227 0.014
0.022 0.92
0.223 0.049
0.023 0.63
0.344 0.022
0.007 0.86
0.260 −0.104 0.197 −0.185
0.052 0.46 0.15 0.15
0.252 −0.094
0.056 0.50
0.262
0.049
0.120 0.035
0.18 0.70
0.098 −0.041 0.529 0.036
0.28 0.65 b0.001 0.76
0.115 0.079 0.387 −0.085 0.298 −0.115
0.20 0.51 0.002 0.51 0.018 0.31
0.401 −0.098
b0.001 0.44
0.301
0.014
Model 1 includes: age, sex, BMI and HOMAir; Model 2 includes: age, sex, BMI, HOMAir, leptin, adiponectin; Model 3 includes: age, sex, BMI, HOMAir, leptin/adiponectin ratio. β: standardized regression coefficient. hs-CRP, HOMAir, leptin, adiponectin, and the leptin/adiponectin ratio are logarithmically transformed.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
4
R.P.F. Dullaart et al. / Clinical Biochemistry xxx (2015) xxx–xxx
Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.clinbiochem.2015.05.001. Conflict of interest statement MAC and JDO are employees of LabCorp. Acknowledgments The GlycA measurements were performed by LabCorp (Raleigh, North Carolina, USA) at no cost. B. Haandrikman, Endocrine Laboratory, University of Groningen, The Netherlands, performed the adipokine assays. The expert technical help of Dr. R. de Vries, MD, PhD was greatly appreciated. References [1] Gornik O, Lauc G. Glycosylation of serum proteins in inflammatory diseases. Dis Markers 2008;25:267–78. [2] Otvos J, Shalaurova I, Wolak-Dinsmore J, Connelly MA, Mackey RH, Stein JH, et al. GlycA: a nuclear magnetic resonance biomarker of systemic inflammation. Clin Chem 2015;61:14–23. [3] Akinkuolie AO, Buring JE, Ridker PM, Mora S. A novel protein glycan biomarker and future cardiovascular disease events. J Am Heart Assoc Sep 23 2014;3(5). http://dx. doi.org/10.1161/JAHA.114.001221 [pii: e001221]. [4] Akinkuolie AO, Pradhan AD, Buring JE, Ridker PM, Mora S. Novel protein glycan sidechain biomarker and risk of incident type 2 diabetes mellitus. Arterioscler Thromb Vasc Biol Apr 23 2015 [pii: ATVBAHA.115.305635. Epub ahead of print]. [5] Arnold JN, Saldova R, Hamid UM, Rudd PM. Evaluation of the serum N-linked glycome for the diagnosis of cancer and chronic inflammation. Proteomics 2008;8: 3284–93.
[6] Dallinga-Thie GM, Dullaart RPF. Do genome-wide association scans provide additional information on the variation of plasma adiponectin concentrations? Atherosclerosis 2010;208:328–9. [7] Blüher M, Mantzoros CS. From leptin to other adipokines in health and disease: facts and expectations at the beginning of the 21st century. Metabolism 2015;64:131–45. [8] Finucane FM, Luan J, Wareham NJ, Sharp SJ, O'Rahilly S, Balkau B, et al. Correlation of the leptin:adiponectin ratio with measures of insulin resistance in non-diabetic individuals. Diabetologia 2009;52:2345–9. [9] Dullaart RP, Kappelle PJ, Dallinga-Thie GM. Carotid intima media thickness is associated with plasma adiponectin but not with the leptin:adiponectin ratio independently of metabolic syndrome. The plasma leptin/adiponectin ratio predicts first cardiovascular event in men: a prospective nested case-control study. Atherosclerosis 2010;211:393–6. [10] Karakas M, Zierer A, Herder C, Baumert J, Meisinger C, Koenig W, et al. Leptin, adiponectin, their ratio and risk of coronary heart disease: results from the MONICA/KORA Augsburg Study 1984–2002. Atherosclerosis 2010;209:220–5. [11] Kappelle PJ, Dullaart RPF, van Beek AP, Hillege HL, Wolffenbuttel BHR. The plasma leptin/adiponectin ratio predicts first cardiovascular event in men: a prospective nested case-control study. Eur J Intern Med 2012;23:755–9. [12] Jeyarajah EJ, Cromwell WC, Otvos JD. Lipoprotein particle analysis by nuclear magnetic resonance spectroscopy. Clin Lab Med 2006;26:847–70. [13] Dullaart RPF, de Vries R, van Tol A, Sluiter WJ. Lower plasma adiponectin is a marker of increased intima-media thickness associated with type 2 diabetes mellitus and with male gender. Eur J Endocrinol 2007;156:387–94. [14] Fischer K, Kettunen J, Würtz P, Haller T, Havulinna AS, Kangas AJ, et al. Biomarker profiling by nuclear magnetic resonance spectroscopy for the prediction of allcause mortality: an observational study of 17,345 persons. PLoS Med 2014;11: e1001606. [15] Emoto M, Nishizawa Y, Maekawa K, Hiura Y, Kanda H, Kawagishi T, et al. Homeostasis model assessment as a clinical index of insulin resistance in type 2 diabetic patients treated with sulfonylureas. Diabetes Care 1999;22:818–22.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Short Communication
GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose tolerance status Robin P.F. Dullaart a,⁎, Eke G. Gruppen a,b, Margery A. Connelly c, James D. Otvos c, Joop D. Lefrandt d a
Department of Endocrinology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands Department of Nephrology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands LabCorp, Raleigh, NC, USA d Department of Vascular Medicine, University of Groningen, University Medical Center Groningen, The Netherlands b c
a r t i c l e
i n f o
Article history: Received 27 March 2015 Received in revised form 1 May 2015 Accepted 2 May 2015 Available online xxxx Keywords: C-reactive protein Glycoproteins Diabetes mellitus Leptin/adiponectin ratio Nuclear magnetic resonance spectroscopy
a b s t r a c t Objectives: Plasma GlycA is a recently developed biomarker whose nuclear magnetic resonance signal originates from glycosylated acute-phase proteins. The aim of our study was to determine potential relationships between GlycA and adiposity, insulin resistance (HOMAir), high sensitive C-reactive protein (hs-CRP), leptin, adiponectin, and the leptin/adiponectin ratio, and to test whether GlycA is elevated in subjects with impaired fasting glucose (IFG) and type 2 diabetes mellitus (T2DM). Design and methods: Plasma GlycA, hs-CRP, leptin, adiponectin, the leptin/adiponectin ratio, and insulin resistance (HOMAir) were measured in 103 fasting subjects (30 with normal fasting glucose, 25 with IFG and 48 with T2DM). Results: In all subjects combined, plasma GlycA was correlated positively with body mass index (BMI), HOMAir, hs-CRP, leptin and the leptin/adiponectin ratio, and inversely with adiponectin (p b 0.05 to p b 0.001). GlycA did not significantly vary according to glucose tolerance category (p = 0.060). GlycA was related positively to the leptin/adiponectin ratio (p = 0.049), independent of BMI (p = 0.056) and HOMAir (p = 0.50). Conclusions: High plasma GlycA reflects a pro-inflammatory state. Adipose tissue-associated inflammatory processes could contribute to increased circulating levels of glycosylated acute-phase proteins. © 2015 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Protein glycosylation, i.e. the enzymatic process whereby a glycan (carbohydrate) moiety is linked to a protein, is influenced by many biological processes including inflammation [1]. Proton nuclear magnetic resonance (NMR) spectroscopy has the ability to detect signals from circulating glycoproteins [2]. Interestingly, the NMRbased glycoprotein biomarker, designated GlycA, which captures both the protein levels and enhanced glycosylation states of the most abundantly expressed acute-phase proteins, has been shown to predict the development of cardiovascular disease (CVD), as well as the progression to type 2 diabetes mellitus (T2DM) in women [3,4]. Inflammatory processes are characterized by enhanced acute phase protein secretion and glycosylation [5]. Accordingly, a positive relationship of GlycA with high-sensitive C-reactive protein (hs-CRP) has been observed [2,3], but insights into the relationships between GlycA and other cardiometabolic risk factors are still limited. Among adipokines, leptin and adiponectin play key roles in adipocyte differentiation and ⁎ Corresponding author at: Department of Endocrinology, University Medical Center Groningen, University of Groningen, P.O. Box 30.001, Groningen 9700 RB, The Netherlands. Fax: +31 503619392. E-mail address: [email protected] (R.P.F. Dullaart).
inflammation [6,7]. The leptin/adiponectin ratio is a determinant of insulin resistance in non-diabetic individuals, highlighting the contribution of adipose tissue dysfunction to the pathogenesis of diminished insulin action [8]. Furthermore, a high leptin/adiponectin ratio may confer increased intima media thickness and predict incident CVD [9–11]. Given that GlycA may predict CVD and progression to T2DM, we hypothesized that GlycA levels may be related to disturbances in pro- and anti-inflammatory adipokines. Therefore, the aim of this study was to interrogate potential relationships of GlycA between and hs-CRP, adiposity, insulin resistance, leptin, adiponectin and the leptin/adiponectin ratio. Second, we tested whether plasma GlycA is elevated in subjects with varying degrees of glucose intolerance. Materials and methods Subjects The medical ethics committee of the University Medical Center Groningen approved the study protocol. Participants with and without T2DM were aged N18 years and were recruited by advertisement in local newspapers. Written informed consent was obtained. T2DM had been diagnosed previously (fasting plasma glucose ≥ 7.0 mmol/L; non-fasting plasma glucose ≥11.1 mmol/L). In non-diabetic subjects,
http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001 0009-9120/© 2015 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
2
R.P.F. Dullaart et al. / Clinical Biochemistry xxx (2015) xxx–xxx
glucose tolerance status was classified as normal fasting glucose (NFG; plasma glucose b5.6 mmol/L) or impaired fasting glucose (IFG; plasma glucose ≥ 5.6 and ≤ 6.9 mmol/L), according to NCEP-ATPIII criteria. T2DM subjects using insulin, current smokers and subjects who used lipid lowering drugs were excluded, as were subjects with a history of CVD, chronic kidney disease, liver function abnormalities or thyroid dysfunction. Subjects using metformin, sulfonylurea or antihypertensive medication were allowed to participate. Blood pressure was measured after 15 min rest at the left arm in sitting position using a sphygmomanometer. Mean arterial pressure was calculated as 1 / 3 × systolic blood pressure + 2 / 3 × diastolic blood pressure. Body mass index (BMI in kg/m2) was calculated as weight divided by height squared. Homeostasis model assessment of insulin resistance (HOMAir) was used to estimate insulin resistance (fasting insulin (mU/L) × fasting glucose (mmol/L) / 22.5).
and the leptin/adiponectin ratio were used. Differences in continuous variables according to glucose tolerance status category were assessed by one-way analysis of variance (ANOVA) with Bonferroni correction for multiple comparisons. Differences in proportions in categorical variables between the glucose tolerance groups were assessed by χ2analysis. The relationship of GlycA and hs-CRP with T2DM was also determined using receiver operating characteristic (ROC) analysis, presented as the area under the curve (AUC) with 95% confidence intervals (CI). Univariate relationships were assessed with Pearson correlation coefficients. Partial correlation coefficients were calculated after adjustment for glucose tolerance category, age and sex. Multivariable linear regression analyses were carried out to disclose the independent contribution of variables to GlycA and hs-CRP. Two-sided p-values b0.05 were considered significant. Results
Laboratory methods Venous blood was collected in EDTA containing tubes. Plasma was stored at −80 °C until analysis. Plasma glucose was measured shortly after blood collection with an APEC glucose analyzer. Total cholesterol and triglycerides were measured by routine enzymatic methods. HDL cholesterol was assayed by a homogeneous enzymatic colorimetric test. NMR spectra were obtained from plasma samples using the NMR Profiler [12]. NMR signal amplitudes originating from the N-acetyl methyl group protons of the N-acetylglucosamine moieties located on the bi-, tri-, and tetra-antennary branches of plasma proteins, predominantly α1-acid glycoprotein, haptoglobin, α1-antitrypsin, α1antichymotrypsin, and transferrin, were used to calculate the GlycA concentration (μmol/L of N-acetyl methyl groups) [2,3]. The intra- and inter-assay coefficients of variation (CV) were 1.9% and 2.6%, respectively [2]. Leptin and total adiponectin were assayed as described [13]. Intra-assay and inter-assay CVs were b 6% and b8%, respectively. Insulin, hs-CRP and glycated hemoglobin (HbA1c) were determined as described [13]. Statistical analysis SPSS (version 20.0, SPSS Inc. Chicago, IL, USA) was used for data analysis. Results are expressed as mean ± SD or as median (interquartile range). Because of skewed distributions, logarithmically transformed values of triglycerides, insulin, HOMAir, hs-CRP, leptin, adiponectin,
One hundred and three subjects participated, of whom 30 were classified with NFG and 25 with IFG (Table 1). Of the 48 subjects with T2DM, 11 used metformin, 9 used sulfonylurea and 15 used both drugs. Antihypertensive medication (mostly angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, diuretics and βblockers, alone or in combination) were used by 23 T2DM subjects. Glucose lowering drugs and antihypertensive medication were not used by non-diabetic subjects. One postmenopausal woman with T2DM, 1 postmenopausal woman with IFG 1, and 1 premenopausal woman with NFD used estrogens. Other medications were not taken. T2DM subjects were older, had higher blood presssure, were more obese and more insulin resistant compared to NFG and IFG subjects (Table 1). There were no gender differences between the groups. Total cholesterol did not differ between the groups, but triglycerides were higher and HDL cholesterol was lower in T2DM than in IFG and NFG subjects. Leptin and the leptin/adiponectin ratio were higher, whereas adiponectin was lower in T2DM compared to IFG and NFG subjects. hs-CRP was higher in T2DM than in IFG subjects. The difference in GlycA between the 3 glucose tolerance groups was not significant. There was also no difference in GlycA between the glucose tolerance groups after age and sex adjustment (p N 0.30). However, the AUC of the ROC curve for GlycA with T2DM vs. no T2DM as the outcome was significant (0.624 (95% CI), 0.515–0.733; p = 0.030). For comparison, the AUC for hs-CRP with T2DM was 0.662 (95% CI, 0.558–0.766; p = 0.005). For GlycA, equal sensitivity and specificity (0.574) were reached
Table 1 Clinical characteristics, plasma glucose, insulin, insulin resistance, plasma lipoproteins, adipokines, high sensitivity C-reactive protein and GlycA in subjects with normal fasting glucose (n = 30), impaired fasting glucose (n = 25) and type 2 diabetes mellitus (n = 48).
Age (years) Sex (men/women) MAP (mm Hg) BMI (kg/m2) Waist circumference (cm) Fasting glucose (mmol/L) HbA1c (mmol/mol) Fasting insulin (mU/L) HOMAir (mU × mmol/L2 × 22.5) Total cholesterol (mmol/L) HDL cholesterol (mmol/L) Triglycerides (mmol/L) Leptin (μg/L) Adiponectin (mg/L) Leptin/adiponectin ratio (μg/mg) GlycA (μmol/L) hs-CRP (mg/L)
Normal fasting glucose (n = 30)
Impaired fasting glucose (n = 25)
Type 2 diabetes mellitus (n = 48)
p-Value*
52 ± 10b 11/19 99.3 ± 15.2b 25.9 ± 4.6c 85.2 ± 13.0c 5.2 ± 0.3c 34 ± 3c 6 (4–9)c 1.4 (1.0–2.0)c 5.7 ± 1.1 1.5 ± 0.4b 1.33 (0.85–2.03)b 6.8 (3.2–22.7)a 22.7 (15.0–48.1)b 0.34 (0.11–0.74)c 373 ± 48 1.81 (0.59–2.92)
56 ± 9 11/14 96.9 ± 10.9c 25.0 ± 3.0c 87.2 ± 10.5c 6.1 ± 0.4c 33 ± 0.3c 6 (5–8)c 1.7 (1.2–2.3)c 5.8 ± 0.8 1.6 ± 0.4c 1.34 (0.92–1.88)b 5.9 (3.3–23.7)a 21.3 (16.2–40.7)b 0.31 (0.13–0.67)c 364 ± 60 0.85 (0.47–1.90)b
59 ± 9 28/20 108.7 ± 10.4 30.7 ± 4.5 106.7 ± 12.6 9.4 ± 2.3 51 ± 8 12 (9–18) 5.7 (3.5–7.7) 5.5 ± 0.9 1.2 ± 0.3 1.94 (1.49–2.54) 14.7 (6.8–31.9) 14.4 (9.5–26.7) 0.97 (0.39–2.13) 395 ± 52 2.19 (1.36–4.20)
0.003 0.15 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.22 b0.001 0.001 0.01 b0.001 b0.001 0.06 0.002
Data in mean ± SD or in median (interquartile range). BMI: body mass index; HbA1c: glycated hemoglobin; HDL: high density lipoproteins; HOMAir: homeostasis model assessment of insulin resistance; hs-CRP: high sensitive C-reactive protein; MAP: mean arterial pressure. ap b 0.05 vs. subjects with type 2 diabetes mellitus; p-value*: p-value by analysis of variance; bp b 0.01 vs. subjects with type 2 diabetes mellitus; cp b 0.001 vs. subjects with type 2 diabetes mellitus.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
R.P.F. Dullaart et al. / Clinical Biochemistry xxx (2015) xxx–xxx
at a value of 383 μg/L, whereas for hs-CRP equal sensitivity and specificity (0.553) were reached at a value of 1.75 mg/L. In all subjects combined, GlycA was correlated positively with BMI, insulin, hs-CRP, HOMAir, leptin and the leptin/adiponectin ratio, and inversely with adiponectin (Supplementary Table 1A). hs-CRP was also correlated positively with BMI, insulin, HOMAir, leptin and the leptin/ adiponectin ratio, and inversely with adiponectin (Supplementary Table 1A). Similar relationships were observed in partial correlation analysis taking account of age, sex, and glucose tolerance category (Supplementary Table 1B). GlycA was correlated positively with hsCRP in the 3 glucose tolerance groups separately (NFG: r = 0.427, p b 0.01; IFG: r = 0.741, p b 0.001; T2DM: r = 0.555, p b 0.001). Neither GlycA, nor hs-CRP were correlated with plasma glucose (Supplementary Tables 1A and B). Multivariable linear regression analysis was performed to determine the extent to which GlycA and hs-CRP were associated with leptin, adiponectin, and the leptin/adiponectin ratio when taking BMI and HOMAir into account (Table 2). In age- and sex-adjusted analyses, GlycA was related to BMI but not significantly to HOMAir (Table 2A Model 1). GlycA was not independently associated with leptin and adiponectin as separate variables (Model 2). Of note, GlycA was positively related to the leptin/adiponectin ratio, independently of BMI and HOMAir (Model 3). hs-CRP was associated with BMI (Table 2B, Model 1), as well as with leptin (Model 2) and the leptin/ adiponectin ratio (Model 3). The strength of the association of GlycA with the leptin/adiponectin ratio was comparable to that of hs-CRP with the leptin/adiponectin ratio (Table 2A Model 3 vs. Table 2B Model 3). When glucose tolerance status, the use of metformin, sulfonylurea and antihypertensive medication were additionally taken into account, the relationship of GlycA with the leptin/adiponectin ratio (cf. Table 2A, Model 3; β = 0.290, p = 0.032), as well as of hsCRP with the leptin/adiponectin ratio (cf. Table 2B, Model 3; β = 0.293, p = 0.019) remained essentially unaltered.
Discussion This study reveals a strong positive relationship between an NMRbased biomarker of inflammatory glycoproteins, designated GlycA, and hs-CRP. Like hs-CRP, GlycA was independently associated with the leptin/adiponectin ratio, a finding which reinforces recent observations
3
demonstrating that this glycoprotein biomarker signal reflects a proinflammatory state [2–4]. Metabolic profiling holds promise for delineating hitherto unappreciated pathways which may conceivably contribute to the development of atherosclerosis [3,14]. Specific glycan structures of acute-phase proteins are enzymatically modified under chronic inflammatory conditions [1,5]. This suggests that GlycA, which represents an integrated plasma glycoprotein biomarker, reflects a pro-inflammatory state beyond associations with the individually captured proteins. In line with this supposition, an NMR-based risk score, including one of the proteins captured in the GlycA NMR signal, namely α-1-acid glycoprotein, may predict all cause and CVD mortality [14]. Well-documented effects of leptin and adiponectin on inflammatory processes [6,7] prompted us to determine the associations of these adipokines with GlycA in comparison with hs-CRP. As shown in the current report, GlycA was related positively to the leptin/ adiponectin ratio, extending expected associations of hs-CRP with this adipokine ratio [10,11]. Of potential relevance, the relationships of GlycA and hs-CRP with the leptin/adiponectin ratio were independent from BMI, a novel finding which agrees with the hypothesis that abnormalities in adipose tissue function could contribute to a proinflammatory state [8]. Several methodological aspects of our study need further discussion. The cross-sectional design of our study does not allow to address the nature of the observed relationships. For this reason, we cannot exclude the possibility of reverse causation. Furthermore, GlycA did not significantly vary according to the degree of glucose intolerance, although ROC analysis indicated that a significant relation of GlycA with the presence of T2DM. In addition, there was no independent association of GlycA with HOMAir, a validated estimate of insulin resistance also in T2DM [15]. In the interpretation of these results it should be considered that glycemic control was adequate in most T2DM subjects. Therefore, the possible effect of more severe hyperglycemia in elevating GlycA warrants further evaluation. Finally, it should be recognized that the enzymatic process of protein glycosylation has to be discerned from protein glycation, a non-enzymatic process which is accelerated under hyperglycemic circumstances. In conclusion, the glycan biomarker, GlycA, is associated positively with the leptin/adiponectin ratio. It is conceivable that adipose tissueassociated inflammatory processes could contribute to increased levels of glycosylated acute-phase proteins.
Table 2 Multivariable linear regression analyses demonstrating independent relationships of GlycA (A) and high sensitive C-reactive protein (hs-CRP) (B) with body mass index (BMI), homeostasis model assessment of insulin resistance (HOMAir), leptin, adiponectin, and the leptin/adiponectin ratio. Model 1 β A GlycA Age Sex (men/women) BMI HOMAir Leptin Adiponectin Leptin/adiponectin ratio B hs-CRP Age Sex (men/women) BMI HOMAir Leptin Adiponectin Leptin/adiponectin ratio
Model 2 p-Value
β
Model 3 p-Value
β
p-Value
0.204 −0.018
0.038 0.86
0.227 0.014
0.022 0.92
0.223 0.049
0.023 0.63
0.344 0.022
0.007 0.86
0.260 −0.104 0.197 −0.185
0.052 0.46 0.15 0.15
0.252 −0.094
0.056 0.50
0.262
0.049
0.120 0.035
0.18 0.70
0.098 −0.041 0.529 0.036
0.28 0.65 b0.001 0.76
0.115 0.079 0.387 −0.085 0.298 −0.115
0.20 0.51 0.002 0.51 0.018 0.31
0.401 −0.098
b0.001 0.44
0.301
0.014
Model 1 includes: age, sex, BMI and HOMAir; Model 2 includes: age, sex, BMI, HOMAir, leptin, adiponectin; Model 3 includes: age, sex, BMI, HOMAir, leptin/adiponectin ratio. β: standardized regression coefficient. hs-CRP, HOMAir, leptin, adiponectin, and the leptin/adiponectin ratio are logarithmically transformed.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
4
R.P.F. Dullaart et al. / Clinical Biochemistry xxx (2015) xxx–xxx
Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.clinbiochem.2015.05.001. Conflict of interest statement MAC and JDO are employees of LabCorp. Acknowledgments The GlycA measurements were performed by LabCorp (Raleigh, North Carolina, USA) at no cost. B. Haandrikman, Endocrine Laboratory, University of Groningen, The Netherlands, performed the adipokine assays. The expert technical help of Dr. R. de Vries, MD, PhD was greatly appreciated. References [1] Gornik O, Lauc G. Glycosylation of serum proteins in inflammatory diseases. Dis Markers 2008;25:267–78. [2] Otvos J, Shalaurova I, Wolak-Dinsmore J, Connelly MA, Mackey RH, Stein JH, et al. GlycA: a nuclear magnetic resonance biomarker of systemic inflammation. Clin Chem 2015;61:14–23. [3] Akinkuolie AO, Buring JE, Ridker PM, Mora S. A novel protein glycan biomarker and future cardiovascular disease events. J Am Heart Assoc Sep 23 2014;3(5). http://dx. doi.org/10.1161/JAHA.114.001221 [pii: e001221]. [4] Akinkuolie AO, Pradhan AD, Buring JE, Ridker PM, Mora S. Novel protein glycan sidechain biomarker and risk of incident type 2 diabetes mellitus. Arterioscler Thromb Vasc Biol Apr 23 2015 [pii: ATVBAHA.115.305635. Epub ahead of print]. [5] Arnold JN, Saldova R, Hamid UM, Rudd PM. Evaluation of the serum N-linked glycome for the diagnosis of cancer and chronic inflammation. Proteomics 2008;8: 3284–93.
[6] Dallinga-Thie GM, Dullaart RPF. Do genome-wide association scans provide additional information on the variation of plasma adiponectin concentrations? Atherosclerosis 2010;208:328–9. [7] Blüher M, Mantzoros CS. From leptin to other adipokines in health and disease: facts and expectations at the beginning of the 21st century. Metabolism 2015;64:131–45. [8] Finucane FM, Luan J, Wareham NJ, Sharp SJ, O'Rahilly S, Balkau B, et al. Correlation of the leptin:adiponectin ratio with measures of insulin resistance in non-diabetic individuals. Diabetologia 2009;52:2345–9. [9] Dullaart RP, Kappelle PJ, Dallinga-Thie GM. Carotid intima media thickness is associated with plasma adiponectin but not with the leptin:adiponectin ratio independently of metabolic syndrome. The plasma leptin/adiponectin ratio predicts first cardiovascular event in men: a prospective nested case-control study. Atherosclerosis 2010;211:393–6. [10] Karakas M, Zierer A, Herder C, Baumert J, Meisinger C, Koenig W, et al. Leptin, adiponectin, their ratio and risk of coronary heart disease: results from the MONICA/KORA Augsburg Study 1984–2002. Atherosclerosis 2010;209:220–5. [11] Kappelle PJ, Dullaart RPF, van Beek AP, Hillege HL, Wolffenbuttel BHR. The plasma leptin/adiponectin ratio predicts first cardiovascular event in men: a prospective nested case-control study. Eur J Intern Med 2012;23:755–9. [12] Jeyarajah EJ, Cromwell WC, Otvos JD. Lipoprotein particle analysis by nuclear magnetic resonance spectroscopy. Clin Lab Med 2006;26:847–70. [13] Dullaart RPF, de Vries R, van Tol A, Sluiter WJ. Lower plasma adiponectin is a marker of increased intima-media thickness associated with type 2 diabetes mellitus and with male gender. Eur J Endocrinol 2007;156:387–94. [14] Fischer K, Kettunen J, Würtz P, Haller T, Havulinna AS, Kangas AJ, et al. Biomarker profiling by nuclear magnetic resonance spectroscopy for the prediction of allcause mortality: an observational study of 17,345 persons. PLoS Med 2014;11: e1001606. [15] Emoto M, Nishizawa Y, Maekawa K, Hiura Y, Kanda H, Kawagishi T, et al. Homeostasis model assessment as a clinical index of insulin resistance in type 2 diabetic patients treated with sulfonylureas. Diabetes Care 1999;22:818–22.
Please cite this article as: Dullaart RPF, et al, GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin/adiponectin ratio than to glucose toleranc..., Clin Biochem (2015), http://dx.doi.org/10.1016/j.clinbiochem.2015.05.001
