ORIGINAL
ARTICLE
Elevation of a Novel Angiogenic Factor, Leucine-Rich␣2-Glycoprotein (LRG1), Is Associated With Arterial Stiffness, Endothelial Dysfunction, and Peripheral Arterial Disease in Patients With Type 2 Diabetes Sharon L. T. Pek, S. Tavintharan, Xiaomeng Wang, Su Chi Lim, Kaing Woon, Lee Ying Yeoh, Xiaowei Ng, Jianjun Liu, and Chee Fang Sum Clinical Research Unit (S.L.T.P., S.T., S.C.L., K.W., X.N., J.L.), Diabetes Centre (S.T., S.C.L., C.F.S.), and Division of Endocrinology (S.T., S.C.L., C.F.S.), Khoo Teck Puat Hospital, Singapore 768828; Metabolic Disease Research Programme (X.W.), Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore 639798; Division of Nephrology (L.Y.Y.), Department of Medicine, Khoo Teck Puat Hospital, Singapore 768828; Institute of Molecular and Cell Biology (X.W.), A*STAR, Singapore 138673; and Institute of Ophthalmology (X.W.), University College London, London EC1V 9EL, United Kingdom Context: Increased arterial stiffness and endothelial dysfunction are associated with peripheral arterial disease (PAD). Leucine-rich-␣2-glycoprotein (LRG1) is a proangiogenic factor involved in regulation of the TGF signaling pathway. Objective: This study in patients with type 2 diabetes mellitus explored the associations of plasma LRG1 with arterial stiffness, endothelial function, and PAD. Design: Based on the ankle brachial index (ABI), patients were classified as having PAD (ABI ⱕ 0.9) or as being borderline abnormal (ABI, 0.91– 0.99) or normal (ABI, 1.00 –1.40). LRG1 was measured by immunoassay; arterial stiffness, by carotid-femoral pulse-wave velocity and augmentation index; and endothelial function, by laser Doppler flowmetry. Results: A total of 2058 patients with type 2 diabetes mellitus were recruited. Mean age (1 SD) was 57.4 (0.2) years. Patients with PAD (n ⫽ 258) had significantly higher LRG1 compared to patients with borderline ABI and patients without PAD (19.00 [13.50] vs 17.35 [13.30] and 15.28 [10.40] g/mL, respectively; P ⬍ .0001). Multiple regression analysis revealed that female gender (P ⬍ .0001), nonChinese ethnicity (P ⬍ .0001), higher waist circumference (P ⫽ .017), lower estimated glomerular filtration rate (P ⬍ .0001), higher urine albumin-creatinine ratio (P ⫽ .009), lower ABI (P ⬍ .0001), higher pulse wave velocity (P ⫽ .040), and poorer endothelium-dependent vasodilation (P ⫽ .007) were independent significant predictors of higher plasma LRG1 levels. A generalized linear model showed that a 1-SD increase in log LRG1 was associated with an odds ratio of 4.072 (95% confidence interval, 1.889 – 8.777; P ⬍ .0001) for prevalence of PAD, after adjustment for traditional risk factors. Conclusions: Higher LRG1 is a significant predictor for arterial stiffness, endothelial function, and PAD. The pathobiological basis and the temporal relationships of these associations need to be explored by further mechanistic and prospective studies to understand the clinical significance of these findings. (J Clin Endocrinol Metab 100: 1586 –1593, 2015)
he prevalence of type 2 diabetes mellitus (T2DM) has reached epidemic levels in both developed and developing countries globally. In Singapore, the prevalence
T
of T2DM in adults reached an alarming 11.3% in 2010 (1). Vascular complications (coronary artery disease, peripheral arterial disease [PAD], stroke, diabetic nephrop-
ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2015 by the Endocrine Society Received October 20, 2014. Accepted January 23, 2015. First Published Online January 30, 2015
Abbreviations: ABI, ankle brachial index; ACE, angiotensin-converting-enzyme; ACR, albumin-creatinine ratio; AI, augmentation index; BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; ENG, endoglin; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LRG1, leucine-rich␣2-glycoprotein; NO, nitric oxide; PAD, peripheral arterial disease; PWV, pulse wave velocity; SBP, systolic blood pressure; T2DM, type 2 diabetes mellitus.
1586
jcem.endojournals.org
J Clin Endocrinol Metab, April 2015, 100(4):1586 –1593
doi: 10.1210/jc.2014-3855
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 October 2015. at 05:06 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2014-3855
athy, neuropathy, and retinopathy) are the main cause of morbidity and mortality in T2DM. A recent report from the American Heart Association suggested that more patients with T2DM, compared to those without, will die from vascular disease (2, 3). Therefore, the importance of prevention and treatment of vascular complications in T2DM cannot be underestimated. PAD occurs in about 15% of adults with T2DM. In both cross-sectional and prospective studies, several risk factors, including ethnicity, gender, age, smoking, and hypertension, have been shown to be associated with PAD (4 –7). PAD is postulated to involve a combination of endothelial dysfunction, lipid disturbance, platelet activation, thrombosis, oxidative stress, vascular smooth muscle cell activation, altered matrix metabolism, remodeling, and inflammation (8). Although there has been much research into the risk factors and mechanisms involved in PAD, the underlying pathophysiology remains unclear. The endothelium has complex functions in regulating vascular homeostasis. It maintains vascular structure integrity, regulates vascular tone, and controls angiogenesis. It was suggested that impairment of endothelium-dependent vasodilation is an early event in the pathogenesis of vascular complications in patients with diabetes (9, 10). Flow-mediated dilatation is decreased in patients with PAD compared to healthy individuals (11) and is a predictor of cardiovascular mortality in these patients (12). Endothelial dysfunction is hypothesized to be associated with increased arterial stiffness. The carotid-to-femoral pulse wave velocity (PWV), a measure of segmental stiffness from the ascending aorta to the femoral artery, has been used to evaluate arterial stiffness in clinical practice (13). Augmentation index (AI), which measures the reflected wave from the peripheral arteries back toward the heart, is also a marker of systemic arterial compliance and stiffness. Individuals with arterial stiffness have reflected a wave returning earlier in systole, leading to increased AI. Increased arterial stiffness, measured by PWV or AI, has been shown to be associated with PAD (14, 15). Leucine-rich-␣2-glycoprotein 1 (LRG1), a novel angiogenic factor, was recently identified to be up-regulated in mouse models of pathogenic ocular vasculature (16). LRG1 is a highly conserved member of the leucine-rich repeat family of proteins, many of which have been found to be involved in protein-protein interaction, signaling, and cell adhesion (17). Increased LRG1 was found in vitreous fluid obtained from patients with proliferative diabetic retinopathy (PDR) compared with non-PDR control patients (with idiopathic epiretinal membranes) (16). Antibody blockade of LRG1 reduced choroidal neovascularization lesion sizes in the mice models, suggesting that LRG1 could be a potential therapeutic target for wet age-
jcem.endojournals.org
1587
related macular degeneration. The aim of the present study is to determine whether plasma LRG1 is differentially expressed in a cross-sectional cohort of T2DM patients with and without PAD and whether LRG1 correlates with measures of arterial stiffness and endothelial function.
Patients and Methods This study was approved by the institution’s Ethics Review Board, and all participants signed informed consent. We consecutively enrolled adults between 21 and 90 years old with T2DM who were seen in our institution’s Diabetes Centre and a community medical center in Singapore from August 2011 to March 2014 as part of an ongoing study (Singapore study of MAcro-angiopathy and microvascular Reactivity in type 2 Diabetes [SMART2D] as reported previously (18). All patients were seen after a 10-hour overnight fast, with venous blood and urine collected. Anthropometric measurements and screening for complications (including retinal assessment by a trained diabetologist or ophthalmologist) were performed. Measurement of the percentage of body fat was performed using a bioelectrical impedance analysis machine (Inbody 330; Biospace Korea). Biochemical analyses were performed in our institution’s Referral Laboratory, which has been accredited by the College of American Pathologists. Serum total cholesterol, high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol levels were measured using an automatized autoanalyzer (Roche Cobas Integra 700; Roche Diagnostics). Glycated hemoglobin (HbA1c) was measured by a point-of-care immunoassay analyzer (DCA Vantage Analyzer; Siemens), a method based on monoclonal antibody agglutination reaction, which meets the National Glycohemoglobin Standardization Program performance standard. Urinary spot albumin and creatinine were measured using commercial assays (Immulite, Siemens). Blood pressure was taken from participants using an automated blood pressure monitor (Dinamap Pro100V2; Criticon). Ankle pressures were measured using a standardized Doppler ultrasonic device (8 MHz, Smartdop TM 20EX, Bidirectional blood flow detector; Hadeco). Both pressure measurements were carried out after a 5-minute rest in the supine position. Hypertension was defined as systolic blood pressure (SBP) ⱖ 140 mm Hg, or diastolic blood pressure (DBP) ⱖ 90 mm Hg, or those who were on antihypertensive medications. Ankle brachial index (ABI) was calculated as the ratio of the higher of the two systolic pressures (from posterior tibial and dorsalis pedis) at the ankle to the higher of the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 October 2015. at 05:06 For personal use only. No other uses without permission. . All rights reserved.
1588
Pek et al
LRG1, Vascular Stiffness, and PAD
right and left brachial artery pressures. PAD was defined as present if the lower ABI was ⱕ 0.9 (19) or if patients had previous amputations. We additionally classified patients with borderline abnormal ABI as 0.91 ⱕ ABI ⱕ 0.99 using the latest American College of Cardiology Foundation/ American Heart Association guidelines (20). Patients with ABI ⬎ 1.4 were excluded from analyses. Assessment for neuropathy was by a neurothesiometer (Horwell Scientific) for vibration and with a 10-g monofilament for light touch. All foot examinations were performed by diabetes nurse educators who received standardized training and accreditation. Neuropathy was present if an abnormal finding in monofilament (inability to detect at least 8 of 10 points on either foot) or neurothesiometer testing of ⱖ 25 V was detected. Estimated GFR (eGFR) was calculated using the Modification of Diet in Renal Disease equation. Plasma LRG1 was measured by commercially available ELISA kits (Immuno-Biological Laboratories). PWV and AI measurements After a minimum of 5 minutes of rest in the supine position, the carotid-femoral PWV was determined using SphygmoCor equipment and software (SphygmoCor) as described previously (21) and expressed as meters per second. AI was calculated as the difference between early and late pressure peaks divided by pulse pressure.
Endothelial function measurements Microvascular endothelial function assessments were performed by the use of laser Doppler flowmetry/imaging (Lisca PIM 3.0; Lisca Development AB) to measure cutaneous perfusion accompanied by iontophoresis of acetylcholine or sodium nitroprusside to assess endothelium-dependent and endothelium-independent vasodilation, respectively, as described previously (22). Vasodilation was presented as percentage change from baseline.
Statistical analyses All analyses were performed using SPSS version 21 (SPSS Inc). Results for continuous variables with normal distribution were presented as mean ⫾ SD; medians and interquartile ranges were presented for continuous but non-normally distributed variables. Binary variables were presented as absolute numbers or percentages. Binary and continuous variables with normal distribution were analyzed using 2 and ANOVA, respectively, whereas nonparametric tests were used for quantitative variables not normally distributed. Bonferroni correction was used for multiple testing. Bivariate associations of LRG1 with clinical variables were analyzed using Spearman’s analyses. Associations and predictors of LRG1 with clinical parameters were analyzed using generalized linear regression. Logistic regression models were used to estimate the odds ratio for PAD and 95% confidence interval (CI) for each risk factor, after adjustment. Skewed data such as LRG1, urinary albumin-creatinine ratio (ACR), and endothelial-dependent and endothelial-independent vasodilation were analyzed after logarithmic transformation. All statis-
J Clin Endocrinol Metab, April 2015, 100(4):1586 –1593
tical tests were two-sided with a level of significance being P ⬍ .05.
Results A total of 2058 T2DM patients were recruited in SMART2D, with mean age of 57.4 ⫾ 0.2 years. The cohort consisted of 49.2% females, and the ethnic distribution was: 51.2% Chinese, 22.4% Malays, 23.1% Indians, and 3.3% of other ethnicity. Eighty-seven participants with missing values of ABI and 44 patients with uncompressible vessels (ABI ⬎ 1.4) were excluded from further analyses. Mean ABI of the remaining 1927 patients was 1.04 ⫾ 0.14, and prevalence of PAD was 13.4%. Characteristics of patients defined by PAD status are summarized in Table 1. In addition to traditional associations of PAD, PWV was significantly higher, and endothelial-dependent vasodilation was significantly poorer in patients with PAD compared to patients with borderline ABI or controls with PAD (P ⬍ .05 for all parameters). Plasma LRG1 was significantly higher in PAD and borderline ABI compared to controls [19.00 (13.50) and 17.35 (13.30) vs 15.28 (10.40) g/ml, P ⬍ .0001, respectively]. [Data are presented as median (interquartile range), where the values of interquatile range is 75th percentile ⫺ 25th percentile) 19.00 (13.50) g/ml is the value for patients with PAD, 17.35 (13.30) g/ml is the value for patients with borderline ABI, 15.28 (10.40) g/ml is the value for patients without PAD.] Bivariate analyses showed that plasma LRG1 was significantly and positively associated with age (rho ⫽ 0.080; P ⫽ .001), body mass index (BMI) (rho ⫽ 0.078; P ⫽ .001), waist circumference (rho ⫽ 0.153; P ⬍ .0001), SBP (rho ⫽ 0.114; P ⬍ .0001), urine ACR (rho ⫽ 0.163; P ⬍ .0001), PWV (rho ⫽ 0.119; P ⬍ .0001), and AI (rho ⫽ 0.136; P ⬍ .0001) and was negatively associated with fasting HDL (rho ⫽ ⫺0.050; P ⫽ .027), eGFR (rho ⫽ ⫺0.217; P ⬍ .0001), endothelialdependent vasodilation (rho ⫽ ⫺0.053; P ⫽ .022), and ABI (rho ⫽ ⫺0.184; P ⬍ .0001) (Table 2). Women had higher LRG1 (17.87 [12.5] g/mL) levels compared to men (14.63 [10.04] g/mL; P ⬍ .0001). Measurement of the percentage body fat performed using a bioelectrical impedance analysis machine showed that visceral fat area ( ⫽ 0.127; P ⬍ .0001) and skeletal muscle mass ( ⫽ ⫺0.111; P ⬍ .0001) were correlated with plasma LRG1 (Supplemental Table 1). Malays and Indians had significantly higher LRG1 compared to Chinese (18.57 [13.05] and 18.30 [12.96] g/mL vs 14.74 [9.87] g/mL, respectively; P ⬍ .0001). Patients on angiotensin-convertingenzyme (ACE) inhibitors, statins, and insulin had significantly higher plasma LRG1, whereas patients on metformin had significantly lower plasma LRG1 (Supple-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 October 2015. at 05:06 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2014-3855
Table 1.
jcem.endojournals.org
1589
Baseline Characteristics of Study Subjects
Characteristics
Non-PAD
Borderline ABI
PAD
n Age, y Gender, % males Ethnicity, % Chinese Malay Indian Others BMI, kg/m2 Waist circumference, cm Total cholesterol, mmol/L HDL cholesterol, mmol/L LDL cholesterol, mmol/L HbA1c, mmol/mol HbA1c, % SBP, mm Hg DBP, mm Hg eGFR, mL/min/1.73 m2 Urine ACR, mg/g PWV, m/s AI, % Endothelium-independent vasodilation, % Endothelium-dependent vasodilation, % ABI LRG1, g/mL Neuropathy, n (%) Retinopathy, n (%) No retinopathy Non-proliferative retinopathy Proliferative retinopathy History of smoking, % Nonsmoker Ex-smoker Smoker On medications, % Metformin Insulin ACE inhibitor Angiotensin II Receptor Blocker Statins Fibrates
1312 56.5 ⫾ 10.5 55.3
357 58.4 ⫾ 11.0 38.9
258 61.5 ⫾ 10.5 41.1
54.2 22.3 20.3 3.3 27.8 ⫾ 5.1 96.6 ⫾ 14.7 4.42 ⫾ 1.05 1.30 ⫾ 0.35 2.75 ⫾ 0.83 62 ⫾ 15 7.76 ⫾ 1.32 136.6 ⫾ 18.5 78.0 ⫾ 9.5 86.40 ⫾ 32.56 21.0 (83.0) 9.63 ⫾ 2.78 25.55 ⫾ 10.55 67 (67.4) 106.8 (126.3) 1.11 ⫾ 0.09 15.28 (10.40) 128 (9.9)
45.9 21.0 30.3 2.8 27.3 ⫾ 5.4 97.1 ⫾ 15.0 4.43 ⫾ 0.95 1.30 ⫾ 0.38 2.75 ⫾ 0.85 61 ⫾ 14 7.75 ⫾ 1.26 140.5 ⫾ 18.3 77.1 ⫾ 9.4 86.36 ⫾ 32.66 22.0 (91.3) 9.62 ⫾ 2.92 28.35 ⫾ 10.60 68.2 (69.1) 97.7 (123.3) 0.94 ⫾ 0.03 17.35 (13.30) 30 (8.5)
46.1 24.8 25.6 3.5 26.7 ⫾ 5.0 95.2 ⫾ 14.2 4.47 ⫾ 0.97 1.26 ⫾ 0.36 2.81 ⫾ 0.87 62 ⫾ 15 7.83 ⫾ 1.40 145.2 ⫾ 21.8 76.9 ⫾ 9.6 74.23 ⫾ 33.50 42.0 (218.0) 10.30 ⫾ 3.29 29.20 ⫾ 9.89 58.2 (71.8) 91.1 (110.4) 0.84 ⫾ 0.20 19.00 (13.50) 37 (14.9)
939 (74.1) 213 (16.9) 115 (9.1)
253 (73.8) 61 (17.8) 29 (8.5)
162 (65.3) 48 (19.4) 38 (15.3)
83.9 7.9 8.3
84.3 5.3 10.4
81.9 8.5 9.7
82.0 27.1 36.1 26.9 81.8 10.2
81.7 26.1 37.4 24.2 79.6 6.5
75.9 33.5 38.3 26.8 82.4 14.1
P Value
ARTICLE
Elevation of a Novel Angiogenic Factor, Leucine-Rich␣2-Glycoprotein (LRG1), Is Associated With Arterial Stiffness, Endothelial Dysfunction, and Peripheral Arterial Disease in Patients With Type 2 Diabetes Sharon L. T. Pek, S. Tavintharan, Xiaomeng Wang, Su Chi Lim, Kaing Woon, Lee Ying Yeoh, Xiaowei Ng, Jianjun Liu, and Chee Fang Sum Clinical Research Unit (S.L.T.P., S.T., S.C.L., K.W., X.N., J.L.), Diabetes Centre (S.T., S.C.L., C.F.S.), and Division of Endocrinology (S.T., S.C.L., C.F.S.), Khoo Teck Puat Hospital, Singapore 768828; Metabolic Disease Research Programme (X.W.), Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore 639798; Division of Nephrology (L.Y.Y.), Department of Medicine, Khoo Teck Puat Hospital, Singapore 768828; Institute of Molecular and Cell Biology (X.W.), A*STAR, Singapore 138673; and Institute of Ophthalmology (X.W.), University College London, London EC1V 9EL, United Kingdom Context: Increased arterial stiffness and endothelial dysfunction are associated with peripheral arterial disease (PAD). Leucine-rich-␣2-glycoprotein (LRG1) is a proangiogenic factor involved in regulation of the TGF signaling pathway. Objective: This study in patients with type 2 diabetes mellitus explored the associations of plasma LRG1 with arterial stiffness, endothelial function, and PAD. Design: Based on the ankle brachial index (ABI), patients were classified as having PAD (ABI ⱕ 0.9) or as being borderline abnormal (ABI, 0.91– 0.99) or normal (ABI, 1.00 –1.40). LRG1 was measured by immunoassay; arterial stiffness, by carotid-femoral pulse-wave velocity and augmentation index; and endothelial function, by laser Doppler flowmetry. Results: A total of 2058 patients with type 2 diabetes mellitus were recruited. Mean age (1 SD) was 57.4 (0.2) years. Patients with PAD (n ⫽ 258) had significantly higher LRG1 compared to patients with borderline ABI and patients without PAD (19.00 [13.50] vs 17.35 [13.30] and 15.28 [10.40] g/mL, respectively; P ⬍ .0001). Multiple regression analysis revealed that female gender (P ⬍ .0001), nonChinese ethnicity (P ⬍ .0001), higher waist circumference (P ⫽ .017), lower estimated glomerular filtration rate (P ⬍ .0001), higher urine albumin-creatinine ratio (P ⫽ .009), lower ABI (P ⬍ .0001), higher pulse wave velocity (P ⫽ .040), and poorer endothelium-dependent vasodilation (P ⫽ .007) were independent significant predictors of higher plasma LRG1 levels. A generalized linear model showed that a 1-SD increase in log LRG1 was associated with an odds ratio of 4.072 (95% confidence interval, 1.889 – 8.777; P ⬍ .0001) for prevalence of PAD, after adjustment for traditional risk factors. Conclusions: Higher LRG1 is a significant predictor for arterial stiffness, endothelial function, and PAD. The pathobiological basis and the temporal relationships of these associations need to be explored by further mechanistic and prospective studies to understand the clinical significance of these findings. (J Clin Endocrinol Metab 100: 1586 –1593, 2015)
he prevalence of type 2 diabetes mellitus (T2DM) has reached epidemic levels in both developed and developing countries globally. In Singapore, the prevalence
T
of T2DM in adults reached an alarming 11.3% in 2010 (1). Vascular complications (coronary artery disease, peripheral arterial disease [PAD], stroke, diabetic nephrop-
ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2015 by the Endocrine Society Received October 20, 2014. Accepted January 23, 2015. First Published Online January 30, 2015
Abbreviations: ABI, ankle brachial index; ACE, angiotensin-converting-enzyme; ACR, albumin-creatinine ratio; AI, augmentation index; BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; ENG, endoglin; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LRG1, leucine-rich␣2-glycoprotein; NO, nitric oxide; PAD, peripheral arterial disease; PWV, pulse wave velocity; SBP, systolic blood pressure; T2DM, type 2 diabetes mellitus.
1586
jcem.endojournals.org
J Clin Endocrinol Metab, April 2015, 100(4):1586 –1593
doi: 10.1210/jc.2014-3855
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 October 2015. at 05:06 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2014-3855
athy, neuropathy, and retinopathy) are the main cause of morbidity and mortality in T2DM. A recent report from the American Heart Association suggested that more patients with T2DM, compared to those without, will die from vascular disease (2, 3). Therefore, the importance of prevention and treatment of vascular complications in T2DM cannot be underestimated. PAD occurs in about 15% of adults with T2DM. In both cross-sectional and prospective studies, several risk factors, including ethnicity, gender, age, smoking, and hypertension, have been shown to be associated with PAD (4 –7). PAD is postulated to involve a combination of endothelial dysfunction, lipid disturbance, platelet activation, thrombosis, oxidative stress, vascular smooth muscle cell activation, altered matrix metabolism, remodeling, and inflammation (8). Although there has been much research into the risk factors and mechanisms involved in PAD, the underlying pathophysiology remains unclear. The endothelium has complex functions in regulating vascular homeostasis. It maintains vascular structure integrity, regulates vascular tone, and controls angiogenesis. It was suggested that impairment of endothelium-dependent vasodilation is an early event in the pathogenesis of vascular complications in patients with diabetes (9, 10). Flow-mediated dilatation is decreased in patients with PAD compared to healthy individuals (11) and is a predictor of cardiovascular mortality in these patients (12). Endothelial dysfunction is hypothesized to be associated with increased arterial stiffness. The carotid-to-femoral pulse wave velocity (PWV), a measure of segmental stiffness from the ascending aorta to the femoral artery, has been used to evaluate arterial stiffness in clinical practice (13). Augmentation index (AI), which measures the reflected wave from the peripheral arteries back toward the heart, is also a marker of systemic arterial compliance and stiffness. Individuals with arterial stiffness have reflected a wave returning earlier in systole, leading to increased AI. Increased arterial stiffness, measured by PWV or AI, has been shown to be associated with PAD (14, 15). Leucine-rich-␣2-glycoprotein 1 (LRG1), a novel angiogenic factor, was recently identified to be up-regulated in mouse models of pathogenic ocular vasculature (16). LRG1 is a highly conserved member of the leucine-rich repeat family of proteins, many of which have been found to be involved in protein-protein interaction, signaling, and cell adhesion (17). Increased LRG1 was found in vitreous fluid obtained from patients with proliferative diabetic retinopathy (PDR) compared with non-PDR control patients (with idiopathic epiretinal membranes) (16). Antibody blockade of LRG1 reduced choroidal neovascularization lesion sizes in the mice models, suggesting that LRG1 could be a potential therapeutic target for wet age-
jcem.endojournals.org
1587
related macular degeneration. The aim of the present study is to determine whether plasma LRG1 is differentially expressed in a cross-sectional cohort of T2DM patients with and without PAD and whether LRG1 correlates with measures of arterial stiffness and endothelial function.
Patients and Methods This study was approved by the institution’s Ethics Review Board, and all participants signed informed consent. We consecutively enrolled adults between 21 and 90 years old with T2DM who were seen in our institution’s Diabetes Centre and a community medical center in Singapore from August 2011 to March 2014 as part of an ongoing study (Singapore study of MAcro-angiopathy and microvascular Reactivity in type 2 Diabetes [SMART2D] as reported previously (18). All patients were seen after a 10-hour overnight fast, with venous blood and urine collected. Anthropometric measurements and screening for complications (including retinal assessment by a trained diabetologist or ophthalmologist) were performed. Measurement of the percentage of body fat was performed using a bioelectrical impedance analysis machine (Inbody 330; Biospace Korea). Biochemical analyses were performed in our institution’s Referral Laboratory, which has been accredited by the College of American Pathologists. Serum total cholesterol, high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol levels were measured using an automatized autoanalyzer (Roche Cobas Integra 700; Roche Diagnostics). Glycated hemoglobin (HbA1c) was measured by a point-of-care immunoassay analyzer (DCA Vantage Analyzer; Siemens), a method based on monoclonal antibody agglutination reaction, which meets the National Glycohemoglobin Standardization Program performance standard. Urinary spot albumin and creatinine were measured using commercial assays (Immulite, Siemens). Blood pressure was taken from participants using an automated blood pressure monitor (Dinamap Pro100V2; Criticon). Ankle pressures were measured using a standardized Doppler ultrasonic device (8 MHz, Smartdop TM 20EX, Bidirectional blood flow detector; Hadeco). Both pressure measurements were carried out after a 5-minute rest in the supine position. Hypertension was defined as systolic blood pressure (SBP) ⱖ 140 mm Hg, or diastolic blood pressure (DBP) ⱖ 90 mm Hg, or those who were on antihypertensive medications. Ankle brachial index (ABI) was calculated as the ratio of the higher of the two systolic pressures (from posterior tibial and dorsalis pedis) at the ankle to the higher of the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 October 2015. at 05:06 For personal use only. No other uses without permission. . All rights reserved.
1588
Pek et al
LRG1, Vascular Stiffness, and PAD
right and left brachial artery pressures. PAD was defined as present if the lower ABI was ⱕ 0.9 (19) or if patients had previous amputations. We additionally classified patients with borderline abnormal ABI as 0.91 ⱕ ABI ⱕ 0.99 using the latest American College of Cardiology Foundation/ American Heart Association guidelines (20). Patients with ABI ⬎ 1.4 were excluded from analyses. Assessment for neuropathy was by a neurothesiometer (Horwell Scientific) for vibration and with a 10-g monofilament for light touch. All foot examinations were performed by diabetes nurse educators who received standardized training and accreditation. Neuropathy was present if an abnormal finding in monofilament (inability to detect at least 8 of 10 points on either foot) or neurothesiometer testing of ⱖ 25 V was detected. Estimated GFR (eGFR) was calculated using the Modification of Diet in Renal Disease equation. Plasma LRG1 was measured by commercially available ELISA kits (Immuno-Biological Laboratories). PWV and AI measurements After a minimum of 5 minutes of rest in the supine position, the carotid-femoral PWV was determined using SphygmoCor equipment and software (SphygmoCor) as described previously (21) and expressed as meters per second. AI was calculated as the difference between early and late pressure peaks divided by pulse pressure.
Endothelial function measurements Microvascular endothelial function assessments were performed by the use of laser Doppler flowmetry/imaging (Lisca PIM 3.0; Lisca Development AB) to measure cutaneous perfusion accompanied by iontophoresis of acetylcholine or sodium nitroprusside to assess endothelium-dependent and endothelium-independent vasodilation, respectively, as described previously (22). Vasodilation was presented as percentage change from baseline.
Statistical analyses All analyses were performed using SPSS version 21 (SPSS Inc). Results for continuous variables with normal distribution were presented as mean ⫾ SD; medians and interquartile ranges were presented for continuous but non-normally distributed variables. Binary variables were presented as absolute numbers or percentages. Binary and continuous variables with normal distribution were analyzed using 2 and ANOVA, respectively, whereas nonparametric tests were used for quantitative variables not normally distributed. Bonferroni correction was used for multiple testing. Bivariate associations of LRG1 with clinical variables were analyzed using Spearman’s analyses. Associations and predictors of LRG1 with clinical parameters were analyzed using generalized linear regression. Logistic regression models were used to estimate the odds ratio for PAD and 95% confidence interval (CI) for each risk factor, after adjustment. Skewed data such as LRG1, urinary albumin-creatinine ratio (ACR), and endothelial-dependent and endothelial-independent vasodilation were analyzed after logarithmic transformation. All statis-
J Clin Endocrinol Metab, April 2015, 100(4):1586 –1593
tical tests were two-sided with a level of significance being P ⬍ .05.
Results A total of 2058 T2DM patients were recruited in SMART2D, with mean age of 57.4 ⫾ 0.2 years. The cohort consisted of 49.2% females, and the ethnic distribution was: 51.2% Chinese, 22.4% Malays, 23.1% Indians, and 3.3% of other ethnicity. Eighty-seven participants with missing values of ABI and 44 patients with uncompressible vessels (ABI ⬎ 1.4) were excluded from further analyses. Mean ABI of the remaining 1927 patients was 1.04 ⫾ 0.14, and prevalence of PAD was 13.4%. Characteristics of patients defined by PAD status are summarized in Table 1. In addition to traditional associations of PAD, PWV was significantly higher, and endothelial-dependent vasodilation was significantly poorer in patients with PAD compared to patients with borderline ABI or controls with PAD (P ⬍ .05 for all parameters). Plasma LRG1 was significantly higher in PAD and borderline ABI compared to controls [19.00 (13.50) and 17.35 (13.30) vs 15.28 (10.40) g/ml, P ⬍ .0001, respectively]. [Data are presented as median (interquartile range), where the values of interquatile range is 75th percentile ⫺ 25th percentile) 19.00 (13.50) g/ml is the value for patients with PAD, 17.35 (13.30) g/ml is the value for patients with borderline ABI, 15.28 (10.40) g/ml is the value for patients without PAD.] Bivariate analyses showed that plasma LRG1 was significantly and positively associated with age (rho ⫽ 0.080; P ⫽ .001), body mass index (BMI) (rho ⫽ 0.078; P ⫽ .001), waist circumference (rho ⫽ 0.153; P ⬍ .0001), SBP (rho ⫽ 0.114; P ⬍ .0001), urine ACR (rho ⫽ 0.163; P ⬍ .0001), PWV (rho ⫽ 0.119; P ⬍ .0001), and AI (rho ⫽ 0.136; P ⬍ .0001) and was negatively associated with fasting HDL (rho ⫽ ⫺0.050; P ⫽ .027), eGFR (rho ⫽ ⫺0.217; P ⬍ .0001), endothelialdependent vasodilation (rho ⫽ ⫺0.053; P ⫽ .022), and ABI (rho ⫽ ⫺0.184; P ⬍ .0001) (Table 2). Women had higher LRG1 (17.87 [12.5] g/mL) levels compared to men (14.63 [10.04] g/mL; P ⬍ .0001). Measurement of the percentage body fat performed using a bioelectrical impedance analysis machine showed that visceral fat area ( ⫽ 0.127; P ⬍ .0001) and skeletal muscle mass ( ⫽ ⫺0.111; P ⬍ .0001) were correlated with plasma LRG1 (Supplemental Table 1). Malays and Indians had significantly higher LRG1 compared to Chinese (18.57 [13.05] and 18.30 [12.96] g/mL vs 14.74 [9.87] g/mL, respectively; P ⬍ .0001). Patients on angiotensin-convertingenzyme (ACE) inhibitors, statins, and insulin had significantly higher plasma LRG1, whereas patients on metformin had significantly lower plasma LRG1 (Supple-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 October 2015. at 05:06 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2014-3855
Table 1.
jcem.endojournals.org
1589
Baseline Characteristics of Study Subjects
Characteristics
Non-PAD
Borderline ABI
PAD
n Age, y Gender, % males Ethnicity, % Chinese Malay Indian Others BMI, kg/m2 Waist circumference, cm Total cholesterol, mmol/L HDL cholesterol, mmol/L LDL cholesterol, mmol/L HbA1c, mmol/mol HbA1c, % SBP, mm Hg DBP, mm Hg eGFR, mL/min/1.73 m2 Urine ACR, mg/g PWV, m/s AI, % Endothelium-independent vasodilation, % Endothelium-dependent vasodilation, % ABI LRG1, g/mL Neuropathy, n (%) Retinopathy, n (%) No retinopathy Non-proliferative retinopathy Proliferative retinopathy History of smoking, % Nonsmoker Ex-smoker Smoker On medications, % Metformin Insulin ACE inhibitor Angiotensin II Receptor Blocker Statins Fibrates
1312 56.5 ⫾ 10.5 55.3
357 58.4 ⫾ 11.0 38.9
258 61.5 ⫾ 10.5 41.1
54.2 22.3 20.3 3.3 27.8 ⫾ 5.1 96.6 ⫾ 14.7 4.42 ⫾ 1.05 1.30 ⫾ 0.35 2.75 ⫾ 0.83 62 ⫾ 15 7.76 ⫾ 1.32 136.6 ⫾ 18.5 78.0 ⫾ 9.5 86.40 ⫾ 32.56 21.0 (83.0) 9.63 ⫾ 2.78 25.55 ⫾ 10.55 67 (67.4) 106.8 (126.3) 1.11 ⫾ 0.09 15.28 (10.40) 128 (9.9)
45.9 21.0 30.3 2.8 27.3 ⫾ 5.4 97.1 ⫾ 15.0 4.43 ⫾ 0.95 1.30 ⫾ 0.38 2.75 ⫾ 0.85 61 ⫾ 14 7.75 ⫾ 1.26 140.5 ⫾ 18.3 77.1 ⫾ 9.4 86.36 ⫾ 32.66 22.0 (91.3) 9.62 ⫾ 2.92 28.35 ⫾ 10.60 68.2 (69.1) 97.7 (123.3) 0.94 ⫾ 0.03 17.35 (13.30) 30 (8.5)
46.1 24.8 25.6 3.5 26.7 ⫾ 5.0 95.2 ⫾ 14.2 4.47 ⫾ 0.97 1.26 ⫾ 0.36 2.81 ⫾ 0.87 62 ⫾ 15 7.83 ⫾ 1.40 145.2 ⫾ 21.8 76.9 ⫾ 9.6 74.23 ⫾ 33.50 42.0 (218.0) 10.30 ⫾ 3.29 29.20 ⫾ 9.89 58.2 (71.8) 91.1 (110.4) 0.84 ⫾ 0.20 19.00 (13.50) 37 (14.9)
939 (74.1) 213 (16.9) 115 (9.1)
253 (73.8) 61 (17.8) 29 (8.5)
162 (65.3) 48 (19.4) 38 (15.3)
83.9 7.9 8.3
84.3 5.3 10.4
81.9 8.5 9.7
82.0 27.1 36.1 26.9 81.8 10.2
81.7 26.1 37.4 24.2 79.6 6.5
75.9 33.5 38.3 26.8 82.4 14.1
P Value
