Arterial stiffness in diabetes mellitus

Atherosclerosis 238 (2015) 370e379

Contents lists available at ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

Review

Arterial stiffness in diabetes mellitus Stuart B. Prenner a, Julio A. Chirinos a, b, * a b

University of Pennsylvania Perelman School of Medicine, 3400 Spruce Street, 6 Penn Tower, Philadelphia, PA 19104, USA Philadelphia VA Medical Center, Division of Cardiology e Suite 8B111, University & Woodland Avenues, Philadelphia, PA 19104, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 July 2014 Received in revised form 17 November 2014 Accepted 12 December 2014 Available online 20 December 2014

Arterial stiffness is an age-related process that is a shared consequence of numerous diseases including diabetes mellitus (DM), and is an independent predictor of mortality both in this population and in the general population. While much has been published about arterial stiffness in patients with DM, a thorough review of the current literature is lacking. Using a systematic literature search strategy, we aimed to summarize our current understanding related to arterial stiffness in DM. We review key studies demonstrating that, among patients with established DM, arterial stiffness is closely related to the progression of complications of DM, including nephropathy, retinopathy, and neuropathy. It is also becoming clear that arterial stiffness can be increased even in pre-diabetic populations with impaired glucose tolerance, and in those with the metabolic syndrome (METS), well before the onset of overt DM. Some data suggests that arterial stiffness can predict the onset of DM. However, future work is needed to further clarify whether large artery stiffness and the pulsatile hemodynamic changes that accompany it are involved in the pathogenesis of DM, and whether interventions targeting arterial stiffness are associated with improved clinical outcomes in DM. We also review of the potential mechanisms of arterial stiffness in DM, with particular emphasis on the role of advanced glycation endproducts (AGEs) and nitric oxide dysregulation, and address potential future directions for research. © 2014 Elsevier Ireland Ltd. All rights reserved.

Keywords: Diabetes mellitus Arterial stiffness

Contents 1. 2. 3.

4.

5. 6.

Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 Human studies of DM and arterial stiffness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 3.1. Arterial stiffness in prediabetic states and the metabolic syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 3.2. Arterial stiffness and type 1 DM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 3.3. Arterial stiffness and type 2 DM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 Arterial stiffness and microvascular/macrovascular changes in DM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 4.1. Nephropathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 4.2. Retinopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 4.3. Autonomic dysfunction/neuropathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 4.4. Cognitive dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376 Advanced glycation end-products, nitric oxide dysregulation, and arterial stiffness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

1. Background * Corresponding author. Philadelphia Veterans Administration Medical Center, Division of Cardiology e Suite 8B111, University & Woodland Avenues, Philadelphia, PA 19104, USA. E-mail address: [email protected] (J.A. Chirinos). http://dx.doi.org/10.1016/j.atherosclerosis.2014.12.023 0021-9150/© 2014 Elsevier Ireland Ltd. All rights reserved.

Arterial stiffness is an age-related progressive process that is a shared consequence of numerous diseases including diabetes

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379

mellitus (DM), hypertension, the metabolic syndrome (METS) and chronic kidney disease (CKD), among others. DM is as an increasingly prevalent disease with well-recognized short and long-term cardiovascular consequences. In this paper, we discuss recent findings from clinical and population-based studies assessing arterial stiffness in diabetic cohorts. There is a particular emphasis on the impact of arterial stiffening on microvascular and macrovascular complications of the disease. We additionally review the pathophysiologic basis of arterial stiffness in DM and discuss the outcomes of interventions targeting these processes. 2. Methods We searched Medline (from 1946 to 2013) and Embase (from 1947 to 2013) databases, using detailed search strategies, as detailed in the online supplement. We developed search methods to capture publications related to arterial stiffness and DM, AGEs, and microvascular complications of DM. This search generated 1700 articles and abstracts which were reviewed in their entirety and discussed by the authors to establish their relevance to this review. When appropriate, we discuss major contributions to the field in text, and use tables to summarize individual studies. This review focuses on conceptual syntheses of data, rather than detailed descriptions of original research. 3. Human studies of DM and arterial stiffness Arterial stiffness represents a subclinical marker of cardiovascular risk. Arterial stiffness has been shown to be an independent risk factor for adverse cardiovascular events and all-cause mortality in the general population [1]. Markers of large artery stiffness such as pulse pressure (PP), pulse wave velocity (PWV), and aortic characteristic impedance (Zc), have been widely studied over the past two decades. With improvements in techniques to noninvasively measure these hemodynamic parameters, the literature relating arterial stiffness and pulsatile hemodynamics to various disease states has grown significantly. Table 1 provides a brief summary of common indices of arterial stiffness and related pulsatile hemodynamics. Most of these indices are based on either: (1) the velocity of pulse transmission time between 2 arterial points (i.e., PWV, which is considered the non-invasive “gold standard” measurement of arterial stiffness) or (2) aortic distensability. In addition, indices of wave reflections (such as augmentation index or reflection magnitude) are commonly reported in studies of arterial stiffness. Augmentation index is a composite measure of wave reflections and arterial stiffness.

with increased arterial stiffness that was intermediate between that of normal controls and patients with type 2 DM [2]. In another study of 232 previously healthy Japanese adults, brachial-ankle PWV was shown to increase linearly with increasing fasting blood glucose and hemoglobin A1C, and to increase stepwise in patients with normal, IFG, and DM respectively [3]. Interestingly, it has been demonstrated that even within the normal glucose range, stepwise increases in fasting glucose from 65 mg/dL to 100 mg/dL were associated with increased arterial stiffness in 697 patients without DM [4]. Most recently, among 872 adults with normal glucose tolerance, insulin resistance was independently associated with increased brachial-ankle PWV [5]. This suggests that even early insulin resistance that develops prior to impairment in glucose tolerance is associated with arterial stiffness. However, the finding of increased arterial stiffness in “prediabetes” has not been universal. A recent large study in 1927 middleaged men and women (Asklepios cohort) suggested that after controlling for age, sex and mean arterial pressure (MAP), IFG was not associated with intrinsic arterial stiffness [6]. Other studies have shown conflicting results. Li et al. showed that in a large study of 4938 patients, subjects with IGT, but not isolated IFG, exhibit increased arterial stiffness compared to normal controls [7]. Accordingly, the relationship between insulin resistance, IGT, and arterial stiffness requires more investigation and longitudinal studies. Increased arterial stiffness has also been associated with the METS. As part of the Bogalusa Heart study, Li et al. examined arterial stiffness in 806 asymptomatic adults between 22 and 44 years of age. Brachial-ankle PWV increased significantly with the increasing number of METS components. Furthermore, the rate of change Table 1 Key definitions related to arterial stiffness. Index

Definition

Arterial pulse wave velocity (PWV)

Velocity with which the pulse wave travels throughout the arterial tree and a measure of arterial stiffness. PWV is proportional to the square root of the incremental elastic modulus of the vessel wall (a measure of the stiffness of the wall material, see below), given constant ratio of wall thickness to vessel radius. Impedance to pulsatile flow imparted by a specific arterial segment. It is a local arterial property influenced by wall stiffness and thickness, but also strongly influenced by vessel size. Change in arterial volume relative to the change in arterial pressure. Fractional (relative) change in arterial volume relative to the change in arterial pressure. In our study, it was assessed via the distensibility coefficient, which is the fractional (relative) change in arterial cross-sectional area relative to the change in arterial pressure. Measure of wall material stiffness. It is defined as the local slope of the incremental change in circumferential stress and the incremental change in circumferential length (strain) of the wall material at the operating range of stress and strain. It is not measured directly in vivo, but can be inferred from measurements of PWV, Zc or compliance when the diameter and wall thickness of the vessel are known. Proximal aortic property that determines the amount of proximal aortic pressure increase for any given flow increase during early systole. It is an important determinant of aortic pulse pressure. Summed compliance of the arterial tree. It is an important determinant of arterial pulse pressure. Ratio of the amplitude of the reflected wave/amplitude of the incident (forward) wave. The augmentation index is defined as the proportion of central pulse pressure due to the late systolic peak, which is in turn attributed to the reflected pulse wave.

Characteristic impedance (Zc)

Compliance Distensibility

3.1. Arterial stiffness in prediabetic states and the metabolic syndrome In this section we summarize the major studies that assessed arterial stiffness in patients with “pre-diabetes” i.e., impaired glucose tolerance (IGT), insulin resistance, as well as patients with the metabolic syndrome. Studies assessing the relationship between arterial stiffness and/or related hemodynamic indices in impaired glucose tolerance, insulin resistance, the metabolic syndrome, and DM are summarized in Table 2. There is recent increasing evidence that, even before the onset of DM, signs of abnormal arterial stiffness are evident in patients with pre-diabetic states including impaired fasting glucose, glucose intolerance, and insulin resistance. Numerous studies have examined the impact of these states on vascular function. The Hoorn study, a large population based-cohort study of nearly 2500 subjects in the Netherlands, showed that after adjusting for age and blood pressure (BP), impaired fasting glucose (IFG) was associated

371

Incremental elastic modulus (Einc)

Ascending aortic Zc

Total arterial compliance Reflection magnitude Augmentation index (AIx)

372

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379

Table 2 Summary of studies on arterial stiffness and indices of wave reflections in hyperglycemia and DM. Author

n¼

Metabolic condition assessed

Study type

Indices measured

Findings

Brooks et al., 2001 [1] Cameron et al., 2003 [2]

173 169

T2DM T2DM

Cross-sectional Cross-sectional

Chirinos et al., 2011 [3]

10,550

T2DM

Cross-sectional

AIx cfPWV, crPWV, afinPWV, CAC AIx

Chirinos et al., 2013 [4]

1927

T2DM, IFG

Cross-sectional

cfPWV

Choi et al., 2010 [5]

663

IFG, IGT, T2DM

Cross-sectional

baPWV

Cockcroft et al., 2005 [6]

2911

T2DM

Prospective

PP

Cruickshank et al., 2002 [7]

571

T2DM and IGT

Prospective

saPWV

Gordin et al., 2007 [8]

35

T1DM

Experimental

crPWV, cfPWV, AIx

Henry et al., 2003 [9]

747

T2DM, IFG

Cross-sectional

Ho et al., 2011 [10]

2188

T2DM, IFG

Cross-sectional

Carotid, Femoral, and Brachial compliance and distensibility baPWV

Jennersjo et al., 2009 [11] Johansen et al., 2011 [12]

414 896

T2DM Healthy

Cross-sectional Prospective

cfPWV cfPWV

Kimoto et al., 2006 [13]

626

T2DM

Cross-sectional

hfPWV, hcPWV, hbPWV, and afPWV

Kimoto et al., 2003 [14]

290

T2DM

Cross-sectional

Li et al., 2005 [15] Li et al., 2012 [16] Lukich et al., 2010 [17]

806 4938 284

METS IGT, T2DM, IFG T2DM, IFG

Cross-sectional Cross-sectional Cross-sectional

hfPWV, hcPWV, hbPWV, and afPWV baPWV baPWV crPWV, AIx

Nakanishi et al., 2003 [18]

2431

METS

Prospective

cfPWV

Ohnishi et al., 2003 [19] Rahman et al., 2008 [20]

232 90

T2DM, IFG T2DM, IGT

Cross-sectional Cross-sectional

baPWV cfPWV, AIx

Safar et al., 2006 [21]

476

METS

Prospective

cfPWV, PP

Schram et al., 2002 [22]

2484

T2DM

Prospective

PP

Schram et al., 2003 [23]

3250

T1DM

Prospective

PP

Schram et al., 2004 [24]

619

T2DM, IFG

Cross-sectional

Shin et al., 2011 [25] Strain et al., 2006 [26]

697 303

Healthy/IFG T2DM

Cross-sectional Cross-sectional

Total SAC, AIx, and cfTT time baPWV cfPWV, crPWV, fdpPWV

Sweitzer et al., 2007 [27]

34

T1DM

Cross-sectional

cfPWV, PP, Zc

Taniwaki et al., 1999 [28] Tomiyama et al., 2006 [29]

556

T2DM

Cross-sectional

cfPWV

AIx was higher in subjects with T2DM compared to controls. Average CAC was lower and all measures of PWV were increased in subjects with T2DM compared to controls. T2DM was independently associated with a lower AIx among British whites, Africans, and Hispanics, but not among AmericaneIndians or Chinese participants. T2DM, but not IFG, was associated with increased PP, aortic stiffness, but not with increased carotid stiffness. Wave reflections arriving at the central aorta were lower in subjects with T2DM. Mean PWV was higher in subjects with T2DM compared to IFG. 30 min post challenge glucose was independent predictor of PWV. During a 4-year follow-up period, PP independently predicted coronary heart disease, stroke, and peripheral vascular events. PWV was increased in subjects with T2DM compared to controls and independently predicted all-cause and cardiovascular mortality over 10.7 year follow up. Baseline AIx was higher in subjects with T1DM, but increased with hyperglycemia in both subjects with T2DM and controls. baPWV increased during acute hyperglycemia only in subjects with T2DM. T2DM was associated with increased stiffness of carotid, femoral and brachial arteries. IFG was associated with increased femoral and brachial artery stiffness only. baPWV was higher in highest HOMA-IR tertile compared to lowest. PWV was increased stepwise in subjects with IFG and T2DM compared to controls. Non-dipping nocturnal BP was associated with increased PWV. Each yearly increase in HbA1c of 0.1% point was associated with a 0.24 m/s higher PWV over 7.1 year follow up. PWV of each arterial region was increased in subjects with T2DM over controls, and increased further in a stepwise manner with advanced stages of CKD. The increase in PWV was greatest in hfPWV and hcPWV. PWV was increased in subjects with T2DM compared to controls in all four segments, greatest in hfPWV and hcPWV. PWV increased with increasing components of METS Subjects with IGT and T2DM, but not IFG, had increased PWV PWV and AIx were stepwise increased in subjects with IFG and T2DM compared to controls. Increasing components of METS was associating with increasing PWV over time over 9 year follow up. PWV increased as fasting sugar levels increased. PWV was higher in subjects with T2DM, but not IGT, compared to controls. No difference in AIx between groups. PWV increased with increasing number of components of METS. During the 6 year follow-up, increase in PWV was accelerated in the group with METS. This was not observed with PP. PP was associated with increased CV mortality in subjects with T2DM, but not in subjects without T2DM, over 8.6 year follow up. PP was independently associated with increased incidence of CVD over 7.4 year follow up. T2DM was associated with decreased total SAC, increased AIx, and decreased cfTT, whereas IFG was not. IFG was associated with increasing PWV in subjects without T2DM. cfPWV alone was increased in subjects with T2DM compared to controls PP and Zc were elevated in subjects with T1DM compared to controls, whereas PWV was not. cfPWV was higher in subjects with T2DM than controls.

2080

METS

Prospective

baPWV

Van Dijk et al., 2001 [30]

140

IFG

Prospective

PP

Webb et al., 2010 [31]

570

T2DM, IFG

Cross-sectional

cfPWV

Wilikinson et al., 2000 [32]

70

T1DM

Cross-sectional

AIx, Ascending Aortic Pressure and aortic systolic pressure-time integral

PWV was higher in subjects with METS than controls. Rate of increase of the PWV was higher in the subjects with persistent METS than with in subjects with regression of METS PP related to all-cause mortality in subjects with IFG in 6.6 year follow up. Subjects with IFG and T2DM had elevated PWV compared to controls. Fasting glucose concentration, 2 h post-challenge glucose and HOMA-IR were independently associated with PWV AIx and systolic pressure-time integral were elevated in subjects with T2DM compared to controls

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379

373

Table 2 (continued ) Author

n¼

Metabolic condition assessed

Study type

Indices measured

Findings

Witte et al., 2009 [33]

1818

IGT, T2DM, IFG

Cross-sectional

cfPWV

Xu et al., 2010 [34]

1249

T2DM, IFG

Cross-sectional

baPWV

Xu et al., 2010 [35]

1274

IFG, IGT

Cross-sectional

baPWV, PP

Yue et al., 2011 [36]

355

T2DM

Cross-sectional

baPWV, circulating EPCs

Zhang et al., 2011 [37]

158

T2DM

Cross-sectional

cfPWV, crPWV, caPWV, AIx

No difference in PWV was found between isolated IFG and normoglycemia. Subjects with IGT, combined IFG/IGT, and T2DM had higher PWV than controls. In normoglycemia and IFG, increasing Hba1c was associated with increasing PWV PWV and PP were increased in IFG and IGT, compared to controls. Compared to subjects with isolated IFG, those with isolated IGT had increased age- and sex-adjusted PWV Subjects with T2DM had increased PWV compared to controls. EPCs were independent predictor of PWV and Hba1c, and were increased in subjects with better glycemic control cfPWV, crPWV, and caPWV were increased in subjects with T2DM, despite lower AIx.

References are listed in the Online Supplement B. afem ¼ aortic-femoral; afin ¼ aortic-finger; AIx ¼ augmentation index; ba ¼ brachial-ankle; BP ¼ blood pressure; CAC ¼ central aortic compliance; ca ¼ carotid-ankle; cf ¼ carotid-femoral; cfTT ¼ carotid femoral transit time; CKD ¼ chronic kidney disease; cr ¼ carotid-radial; CV ¼ cardiovascular; CVD ¼ cardiovascular disease; EPC ¼ endothelial progenitor cells; fdp ¼ femoral-dorsalis pedis; hb ¼ heart-brachial; HbA1c ¼ hemoglobin a1c; hc ¼ heart-carotid; hf ¼ heart-femoral; HOMAIR ¼ homeostatic model assessment-insulin resistance; IFG ¼ impaired fasting glucose; IGT ¼ impaired glucose tolerance; METS ¼ metabolic syndrome; PP ¼ pulse pressure; PWV ¼ pulse wave velocity; sa ¼ subclavian-aortic; SAC ¼ systemic arterial compliance; T2DM ¼ type 2 diabetes mellitus; T1DM ¼ type 1 diabetes mellitus; Zc ¼ characteristic impedance.

(slope) of PWV with age increased as the number of METS components increased, suggesting that the METS is associated with accelerated age-related changes in arterial stiffness [8]. Safar et al. classified 476 individuals on the basis of number of features of the METS. Carotid-femoral PWV was higher in the group with multiple risk factors, and during 6-year follow-up, the increase in PWV was highest in the group with multiple risk factors, even after adjusting for age and MAP [9]. In a longitudinal study, Tomiyama et al. followed 2080 Japanese men for 3 years. They noted that PWV was higher in patients with the METS even after adjusting for BP. Similar to prior studies, they noted that the rate of increase over time was greatest in the group with METS, but interestingly, that regression of METS was seen to slow the increase in arterial stiffness [10]. This was also supported in the work by Nakanishi et al., which showed that additional features of METS conferred stepwise increase in carotid-femoral PVW over time [11]. Interestingly, among all the METS risk factors, abnormal glucose metabolism may be the key factor driving arterial stiffness. A recent paper by Pietri et al. studied 500 patients with METS and various degrees of abnormal glucose metabolism. Patients underwent carotid-femoral PWV as well as albumin to creatinine ratio (ACR) testing. It was shown that PWV and ACR increased stepwise with increasing degrees of abnormal glucose metabolism, from normal, to IFG, IGT, and Type 2 DM. Furthermore, both ACR and PWV were independently associated with degree of glucose metabolism dysregulation [12]. This finding is important because it suggests that arterial stiffness is not uniformly distributed among patients with METS, and that abnormal glucose metabolism is a key correlate of arterial stiffness in this population. 3.2. Arterial stiffness and type 1 DM While this review is largely dedicated to type 2 DM, studies investigating arterial stiffness in patients with type 1 DM are particularly important, given that these patients often do not demonstrate multiple metabolic comorbidities, such that the impact of hyperglycemia and the diabetic state can be studied with less confounding. Sweitzer et al. measured aortic PWV and pulsatile hemodynamics in 17 patients with type 1 DM (average length of DM was 15 years) and compared them to age- and sex-matched controls. Compared to controls without DM, patients with type 1 DM had elevated central and peripheral PP, despite having similar

mean BP. Additionally, while PWV was not different between the two groups, aortic characteristic impedance (Zc) was increased in the group with DM, and in particular, in patients with albuminuria [13]. Other studies have shown elevated PWV in patients with type 1 DM, while some have only shown this to be true in men [14,15]. The influence of acute hyperglycemia on arterial stiffness was investigated by Gordin et al. Brachial and aortic PWV and augmentation index (AIx) were measured in 22 male patients with type I DM and healthy controls, before and after a 2 h hyperglycemic clamp. Compared to healthy controls, men with type 1 DM had a higher AIx at baseline. While acute hyperglycemia increased AIx in all groups, carotid-radial PWV increased only in patients with type 1 DM [16]. It should be noted that AIx is not a direct index of arterial stiffness, but is rather influenced by wave reflections, arterial stiffness and other factors. Therefore, further studies are required to better assess hemodynamic responses to hyperglycemia, including changes in wave reflections, large artery stiffness and muscular artery stiffness. 3.3. Arterial stiffness and type 2 DM In one study of 169 patient with type 2 DM and non-diabetic controls, patients with type 2 DM had, on average, PWV values similar to 15-year older non-diabetic peers, when measured at carotid-femoral, carotid-radial, and aortic-finger segments [17]. At any level of systolic BP, aortic PWV was greater in patients with type 2 DM than in non-diabetic controls [18,19]. Kimoto et al. measured PWV in the heart-carotid, heart-brachial, heart-femoral, and femoral-ankle segments in patients with type 2 DM and healthy controls. Patients with DM had greater PWV than the healthy subjects in all four arterial segments, but the effect of DM and age were greater in the heart-carotid and heart-femoral segments (large arteries) than in the heart-brachial and femoral-ankle segments (medium sized muscular arteries). DM was significantly associated only with PWV of the central arteries-heart-carotid and heart-femoral segments [20]. Similarly, in a large population of middle-aged subjects, type 2 DM was associated with increased aortic stiffening, but not with increased carotid stiffness [6]. Accordingly, we will focus primarily on studies of central arterial stiffness in subjects with DM that utilize PWV. Initial work into arterial pathology in subjects with type 2 DM aimed to determine the relationship between stiffness and wall

374

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379

Table 3 Summary of studies on arterial stiffness and target organ dysfunction in diabetes. Metabolic condition assessed

Study type

Indices measured

Findings

244

T2DM

Cross-sectional

90

T2DM

Cross-sectional

cfPWV, peripheral and carotid PP baPWV, hcPWV

Bouchi et al., 2011 [3] Cardoso et al., 2009 [4]

461 482

T2DM T2DM

Prospective Cross-sectional

cfPWV cfPWV, crPWV

Chang et al., 2012 [5]

490

T2DM

Cross-sectional

baPWV

Choi et al., 2010 [6] Gouliopoulos et al., 2013 [7] Ha et al., 2012 [8] Ishimura et al., 2007 [9] Kakiba et al., 2004 [10] Kim et al., 2012 [11]

673 202

T2DM T2DM

Cross-sectional Cross-sectional

baPWV cfPWV, AIx

692 167 56 494

T2DM T2DM T2DM T2DM

Cross-sectional Cross Cross-sectional Cross-sectional

Kimoto et al., 2006 [12]

626

T2DM

Cross-sectional

baPWV cfPWV baPWV hfPWV, hcPWV, haPWV, cbPWV, baPWV, faPWV, AIx hfPWV, hcPWV, hbPWV, faPWV

PWV, peripheral and central PP were higher in subjects with T2DM. CKD was independently associated with increased PWV but not PP. baPWV, but not hcPWV was directly related to albuminuria, decreased pulse variation, peripheral neuropathy, and retinopathy. PWV was associated with incident albuminuria and decline in GFR Retinopathy, nephropathy, and neuropathy were independently associated with cfPWV but not crPWV. baPWV increased with worsening CKD stage, but not with worsening albuminuria. Albuminuria was associated with increased PWV. Subjects with retinopathy had increased PWV and AIx compared to those without retinopathy. Subjects with peripheral neuropathy had increased PWV. PWV independently correlated with albuminuria. PWV was increased in subjects with nephropathy. Retinopathy was associated with hfPWV, hcPWV, haPWV, baPWV, but not cbPWV, faPWV or AIx.

Knudsen et al., 2002 [13]

80

T2DM

Cross-sectional

PP on 24-h Monitoring

Meyer et al., 2004 [14]

90

T2DM

Cross-sectional

SAC, cfPWV, fdpPWV

Ogawa et al., 2005 [15] Ogawa et al., 2008 [16] Rema et al., 2004 [17]

1004 201 600

T2DM T2DM T2DM

Cross-sectional Cross-sectional Cross-sectional

baPWV AIx, baPWV AIx

Smith et al., 2005 [18]

134

T2DM

Cross-sectional

cfPWV

Suh et al., 2010 [19] Tanaka et al., 2006 [20]

100 940

T2DM T2DM

Cross-sectional Cross-sectional

baPWV cfPWV

Tanaka et al., 2012 [21]

1099

T2DM

Cross-sectional

baPWV

Tanaka et al., 2013 [22]

689

T2DM

Cross-sectional

baPWV

Theilade et al., 2013 [23]

727

T1DM

Cross-sectional

cfPWV

Yokoyama et al., 2003 [24] Yokoyama et al., 2004 [25] Yokoyama et al., 2007 [26] Yun et al., 2011 [27]

346

T2DM

Cross-sectional

baPWV

306

T2DM

Cross-sectional

baPWV

294

T2DM

Cross-sectional

baPWV, PP

PWV was higher in subjects with T2DM than in controls, and increased as GFR declined. Increase in PWV was greatest in hf and hc segments. After adjustment, decreased GFR was independently associated with increasing hfPWV. Subjects with higher grades of retinopathy had higher PP and blunted diurnal BP variation compared to subjects with lower grades. Similar results were seen in subjects with and without nephropathy. Subjects with diabetes had higher PWV and reduced SAC compared to controls. PWV correlated with worsening autonomic function. Retinopathy was associated with increasing baPWV. PWV, but not AIx, was associated with retinopathy. AIx was higher in subjects with retinopathy, and was associated with retinopathy even after adjusting for microalbuminuria. Subjects with elevated albumin to creatinine ratio had significantly increased PWV. GFR was also associated with increased PWV. A higher baPWV was associated with decreased nerve conduction. SBP and PP were greater in subjects with albuminuria. Increased albuminuria was independently associated with greater PP. Stepwise increases were observed in SBP and PP but not in DBP, in both microalbuminuric and albuminuric subjects. SBP and PP were positively and DBP was negatively associated with PWV. PWV was higher in subject with retinopathy compared to those without. baPWV was higher in subjects with retinopathy, and increased with severity of retinopathy. cfPWV was associated with albuminuria, retinopathy, and autonomic neuropathy. PWV was higher in diabetic subjects with nephropathy than in those without. Albuminuria was significantly associated with baPWV PWV was higher in diabetic subjects with microalbuminuria than those without. PP and PWV were independently associated with neuropathy.

605

T2DM

Cross-sectional

baPWV

Increasing PWV was associated with diabetic retinopathy.

Author

Aoun et al., 2001 [1] Aso et al., 2003 [2]

n¼

References are listed in the Online Supplement B. AIx ¼ augmentation index; ba ¼ brachial-ankle; BP ¼ blood pressure; cb ¼ carotid-brachial; cf ¼ carotid-femoral; CKD ¼ chronic kidney disease; cr ¼ carotid-radial; DBP ¼ diastolic blood pressure; fa ¼ femoral-ankle; fdp ¼ femoral-dorsalis pedis; GFR ¼ glomerular filtration rate; ha ¼ heart-ankle; hb ¼ heart-brachial; hc ¼ heart-carotid; hf ¼ heart-femoral; PP ¼ pulse pressure; PWV ¼ pulse wave velocity; SAC ¼ systemic arterial compliance; SBP ¼ systolic blood pressure; T2DM ¼ type 2 diabetes mellitus; T1DM ¼ type 1 diabetes mellitus.

thickening. In a 1999 study, Taniwaki et al. compared carotid intimal-medial thickness (IMT) and carotid-femoral PWV between populations with and without DM. They showed that there was a positive correlation between measures of arterial thickness and stiffness, and that both were increased in patients with DM. They also showed that PWV increased with age, and that this slope was steeper in patients with type 2 DM than in controls [21]. These findings were corroborated by a similar subsequent study that also demonstrated that changes in thickness and stiffness were observed in patients even with short duration of DM (<5 years), suggesting that many of the vascular changes occur in the prediabetic state or early in the course of DM [22].

Pulse pressure, a rough surrogate of arterial stiffness, has been broadly studied in subjects with type 2 DM. Similar to carotidfemoral PWV, PP has been shown to predict cardiovascular (CV) events in patients with type 2 DM [23,24]. In a longitudinal study, van Dijk et al. showed that in subjects with a history of impaired fasting glucose, elevated brachial artery PP was independently associated with mortality risk, a finding confirmed in subjects with type 2 DM as well [25,26]. The basis for an increased PP in type 2 DM has recently been investigated utilizing comprehensive assessments of arterial hemodynamics in the Asklepios study [6]. It was shown that this was due to increased aortic stiffness and a lower total arterial compliance (most of which is normally provided

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379

by the aorta), rather than increased wave reflections. The latter finding was consistent with a previous study which included over 10,000 subjects from large population-based cohorts around the world and demonstrated that type 2 DM was associated with a reduced AIx among Caucasians, African black adults, and Andean Hispanic adults [27]. DM may lead to preferential increase in large artery stiffness, which may reduce the impedance mismatch with more distal muscular arteries, resulting in a lower reflection magnitude. The impedance matching at the interface between the aorta and medium-sized muscular arteries may also result in increased penetration of pulsatile energy to distal vascular beds, which may contribute to the microvascular complications of DM (discussed below). It is important to note that an independent relationship between arterial stiffness and DM has not been consistently shown across all studies. As noted by Cecelja et al., DM was independently associated with arterial stiffness in roughly 52% of the reported studies, compared to 90% for BP and age. Furthermore, DM accounted for only 5% of the variation in PWV [28]. It is true that DM is likely not a main determinant of arterial stiffness particularly in older, hypertensive patients. However, the independent effects of DM on arterial stiffness are likely obscured to some extent by the association of hypertension with DM. When patients are matched for age and BP, PWV is increased in patients with DM compared to controls without DM, although only in women [29]. Of note, in the Cecelja et al. review, only carotid-femoral PWV was used, narrowing the number of studies that were included. 4. Arterial stiffness and microvascular/macrovascular changes in DM This section reviews current studies examining the relationship between arterial stiffness and target organ dysfunction in DM, specifically: nephropathy, retinopathy, and neuropathy/autonomic dysfunction. Additionally, association between stiffness and cognitive dysfunction and stroke are discussed as well. Studies assessing the relationship between arterial stiffness and/or related hemodynamic indices and target organ dysfunction in type 2 DM are summarized in Table 3. 4.1. Nephropathy DM is a well-known cause of chronic kidney disease (CKD), and along with hypertension, is currently one of the most common causes of progression to end stage renal disease. The development of albuminuria is a poor prognostic sign in populations with and without DM alike, and is an independent predictor of all-cause and cardiovascular mortality [30]. Mild albuminuria may precede the development of renal dysfunction by several years [31]. Numerous studies have assessed the association between arterial stiffness and diabetic nephropathy. Yokoyama et al. studied 306 patients with type 2 DM, approximately one third of whom had microalbuminuria. Brachial-ankle PWV was higher in patients with microalbuminuria than those without. Even after adjustment for conventional cardiovascular risk factors, albuminuria was significantly associated with PWV [32]. These findings were further supported by another study by the same group, in which brachialankle PWV was shown to be stepwise elevated in patients with DM with incipient and clinical nephropathy [33]. One criticism of this study was that the authors did not measure carotid-femoral PWV, which is considered to be the non-invasive gold standard for assessing large artery stiffness. In a subsequent study, Smith et al. measured carotid-femoral PWV in subjects with type 2 DM with normal renal function, but with various degrees of albuminuria. The study found that patients

375

with type 2 DM and elevated albumin to creatinine ratio (ACR) demonstrated increased aortic stiffness compared with those with normal ACR. Furthermore, carotid-femoral PWV was higher in patients within the cohort with lower glomerular filtration rate (GFR). Additionally, the authors estimated that a 10-year increase in the duration of DM affects the PWV to the same extent as a 10 mmHgincrease in PP or a 6.5 year-increase in age [34]. While this study failed to demonstrate that differences in PWV and albuminuria persist after adjusting for BP, other studies have. Ishimura et al. showed a direct relationship between PWV and albuminuria that persisted even after adjusting for BP [35]. It has also been shown that in patients with DM, a decline in GFR over time is associated with increased aortic stiffness [36]. In a longitudinal study, Bouchi et al. showed that during a 6 year follow-up period, aortic stiffness was associated with incident albuminuria and the rate of decline in GFR among 461 patients with type 2 DM [37]. Kimoto et al. examined the relationship between CKD and arterial stiffness among 434 subjects with type 2 DM, compared to 192 healthy controls. PWV was measured in 4 segments (heartfemoral, heart-carotid, heart-brachial, and femoral ankle). PWV of each region was increased in patients with DM compared to healthy controls, but more importantly, increased stepwise with advancing stages of CKD. After adjustment for confounders, decreasing GFR was independently associated with increased PWV of the heartfemoral segment only, suggesting the large artery stiffness, rather than muscular artery stiffness, is associated with impaired GFR [20]. The selective association between a reduced GFR and reduced large artery stiffness, rather than muscular artery stiffness is supported by other studies [36]. Interestingly, aortic PWV, but not carotid-radial or femoral-tibial PWV, predicted mortality in ESRD patients [38]. Multiple studies have also shown an independent relationship between high PP and onset of microalbuminuria. Pedrinelli et al. showed that, among 211 men with untreated hypertension, microalbuminuria was correlated with PP, with increasing quintiles of PP being associated with stepwise increase in urinary albumin excretion [39]. Interestingly it was noted that the association was non-linear, suggesting a threshold above which the hemodynamic effects of elevated PP exerts a renal effect. Similar findings have been shown in other studies [40,41]. It has been suggested that elevated PP may result in excessive penetration of pulsatile energy to the kidney, ultimately damaging the renal microvasculature and leading to wall hypertrophy, as seen in renal biopsies of mildly proteinuric hypertensive patients [42]. 4.2. Retinopathy Arterial stiffness has been shown to be associated with the development of diabetic retinopathy, an independent predictor of mortality in the populations with DM [43,44]. Most recently, Kim et al. 2012 showed that in 494 subjects with type 2 DM, quartiles of heart-femoral PWV were independently associated with diabetic retinopathy [45]. This has been confirmed in several other studies [46e48]. Furthermore, higher brachial-ankle PWV was shown to be stepwise associated with worsening stages of retinopathy [49]. Rema et al. measured aortic AIx among 600 subjects with type 2 DM, approximately 20% of whom had retinopathy. AIx was shown to have a significant positive association with diabetic retinopathy, even after adjusting for age and microalbuminuria [50]. 4.3. Autonomic dysfunction/neuropathy Neural dysfunction is another common manifestation of DM, its prevalence increases with duration of DM, with rates as high as 70% in subjects with long term DM. Neural involvement can manifest

376

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379

either as autonomic (central) or peripheral neuropathy. Autonomic dysfunction, like other microvascular complications of DM, is also an independent predictor of cardiovascular mortality [49,51,52]. Autonomic dysfunction has been associated with worse outcomes in subjects with DM, and has also been linked to other target organ dysfunction, including retinopathy and nephropathy [53e55]. Meyer et al. assessed autonomic dysfunction in 45 type 2 DM and healthy controls. They demonstrated that central PWV correlated with autonomic dysfunction, and hypothesized that autonomic dysfunction leads to blunting of the circadian variation of BP, leading to overall increase in 24-h BP load. They also postulated that a loss of nocturnal dip in heart rate as occurs in autonomic neuropathy, might lead to accelerated fatigue of elastic stretch fibers and thus increased arterial stiffness, by either increasing the number of stretch cycles or by not allowing sufficient relaxation time for large arteries between ventricular contractions [56]. Several other studies have shown associations between elevated PWV and autonomic dysfunction [57]. Even in healthy subjects, elevated baseline HR has been linked to arterial stiffness [58]. Arterial stiffness is associated with peripheral neuropathy among subjects with DM. In a study of 692 type 2 DM, Ha et al. showed that PWV was significantly correlated with peripheral neuropathy as determined with monofilament testing [59]. Other studies that have directly measured nerve conduction have found similar results [60]. Yokoyama et al. assessed diabetic neuropathy as determined by presence of symptoms, absent ankle reflexes, HR variability, and vibratory perception. They showed that PWV was an independent determinant of neuropathy in subjects with type 2 DM [61]. Cardoso et al. conducted a study assessing the correlates of microvascular complications in a large cohort of 482 subjects with type 2 DM without PAD. Subjects underwent carotid-femoral (aortic) and carotid-radial (muscular conduit artery) PWV measurement. As was seen in other studies, no DM-related variables were associated with peripheral arterial stiffness. Retinopathy and nephropathy were, however, independently associated with carotid-femoral PWV, even after adjustment for potential confounders. Increased carotid-femoral PWV was associated with 24h PP [62]. Other studies have found a striking stepwise increase in PP as the degree of albuminuria increased [63]. Knudsen et al. showed that subjects with diabetic retinopathy and albuminuria were found to have a higher PP and blunted diurnal BP variation compared to their diabetic without microvascular complications. The authors suggested that autonomic dysfunction, a known complication of DM, may underlie these changes [64]. However, it is possible that the relationship is bidirectional, with increased PP contributing to the presence of microvascular complications. Overall, available data suggests that large artery stiffness, rather than muscular conduit artery stiffness, is correlated with development of microvascular target organ damage, including retinopathy, nephropathy, and neuropathy. It has been suggested that stiffening of the aorta (which is normally much less stiff than medium-sized muscular arteries), accompanied by little change in medium-sized muscular artery stiffness, leads to impedance “matching” between the aorta and muscular arteries, reducing wave reflections at the aortic-muscular arterial interfaces and increasing transmission of a wider, potentially harmful, forward pulsatile pressure wave to microcirculation. 4.4. Cognitive dysfunction Arterial stiffness is also associated with cognitive changes and white matter lesions among patients with DM. Several recent studies have shown that despite lower BP and cholesterol levels, PWV was higher in patients with type 2 DM than controls, and was

independently associated with larger asymptomatic white matter lesions on MRI [65]. In a study focusing on stroke in patients with type 2 DM, those with symptomatic cerebral infarction had higher brachial-ankle PWV than those without stroke, and this remained significant after controlling for other risk factors [66]. Interestingly, compared to controls, patients with type 2 DM showed evidence of cognitive impairment on neuropsychological testing, and a significant relationship was observed between arterial stiffness and cognitive status [67]. 5. Advanced glycation end-products, nitric oxide dysregulation, and arterial stiffness The remainder of the paper will focus on potential mechanisms of arterial stiffness in DM, with particular emphasis on the role of advanced glycation endproducts (AGEs) and nitric oxide (NO) dysregulation. Several studies have implicated AGEs in the acceleration of agerelated vascular changes and development of cardiovascular events in both diabetic and non-diabetic populations [68]. AGE formation consists of several reversible and irreversible steps that culminate in the detrimental cross-linking of collagen molecules within the arterial vessel wall. Most basically, various sugar moieties (glucose, fructose, glycolytic adductions) interact with the free amino acid residues of proteins, in a reaction described by Louis Maillard in the early 20th century [68e71]. This initial reaction leads to the formation of two intermediate structures e Schiff base and Amadori products, and this reaction is enhanced in the carbonyl-enriched states of DM and end stage renal disease (ESRD) [72]. In the setting of hyperglycemia, Amadori products are formed and unformed rapidly, and are in equilibrium with the blood glucose concentration [73]. Of note, a clinically familiar and relevant Amadori product is glycated hemoglobin (HbA1c), which reflects intermediate-term glycemic levels (approximately over the previous 12 weeks), and is now used for diagnosis of DM [74e76]. These reversible Amadori products undergo slow rearrangements, ultimately leading to the irreversible formation of AGEs [68]. AGE deposition can occur in the arterial wall, largely on collagen, causing pathologic cross-linking [77]. In contrast to crosslinks in normal collagen, which occurs only at two discrete sites at the N-terminal and C-terminal ends of the molecule, AGEs form cross-links throughout the collagen molecule [73]. Additionally, AGE-mediated cross-links may confer a high resistance to enzymatic proteolysis and a decrease in degradation rate, which may contribute to an increased collagen content in arterial walls, characteristic of aging, and accelerated in conditions such as DM [78]. A positive relation between carotid-femoral PWV and collagen crosslinking has been demonstrated [79]. Furthermore, levels of specific AGEs in aortic tissue correlate with aortic stiffness in subjects both with and without DM [80]. In addition to promoting arterial stiffness through AGE formation and collagen cross-linking, insulin resistance contributes to endothelial dysfunction through interfering with NO-mediated vasodilatation. NO has vasodilatory, anti-platelet, anti-inflammatory, and antioxidant properties [81]. In the insulin-resistant state, NO synthase activation is impaired and superoxide production is increased. The combination of both of these factors ultimately leads to reduction in NO bioavailability [82,83] Interestingly, there is emerging evidence that prior to eventual depletion of NO, early insulin resistance may in fact be associated with increased NO levels and low to normal arterial stiffness. Compared to controls, normotensive patients with early insulin resistance showed reduced baseline arterial stiffness indices, yet infusion of an NO synthase inhibitor resulted in greater increase arterial stiffness [84]. Studies in patients with impaired glucose regulation as well as type

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379

2 DM have both shown reduced vasodilation in response to acetylcholine administration, the gold standard for diagnosing endothelial dysfunction [85,86]. In diabetic subjects with microvascular disease compared to those without, basal NO levels were reduced, and NO levels further decreased with increasing severity of microvascular disease [84]. Thus, while initial increase in NO levels may be compensatory with early insulin resistance, there is baseline abnormal response to vasodilation agents, and ultimately NO dysregulation may lead to arterial stiffness and herald the onset of microvascular changes [87]. Given the multiple pathways through which glucose dysregulation may adversely affect arterial stiffness, recent studies have focused on serum levels of AGEs as predictors of poor CV outcomes, and more relevant to this discussion, as determinants of arterial stiffness. Broadly, elevated serum AGE levels have been linked to all cause- and CV-mortality in women but not in men, independent of traditional CV risk factors, in subjects with type 2 DM [88]. In subjects with type 1 DM, higher plasma levels of AGEs were independently associated with incident CV disease and all-cause mortality [89]. Higher soluble receptor for advanced glycation end products (sRAGE) was also associated with increased incident fatal and nonfatal CV disease and all-cause mortality in subjects with type 1 DM [90]. Early work demonstrated that elevated serum and urinary AGEs were significantly associated with PP, an indirect marker of large artery stiffness [91]. More recently, Semba et al. demonstrated that in a communitydwelling population, higher serum AGE levels were significantly associated with carotid-femoral PWV even after adjusting for age, BP, and the presence of DM [92]. However, this has not been consistently replicated. Won et al. showed that while serum AGE levels could predict presence of obstructive coronary artery (CAD) disease in patients with DM, they failed to show an association between serum AGE levels and brachial-ankle PWV. Interestingly, they also failed to demonstrate any difference between arterial stiffness in the patients with DM compared to controls. It should be noted that the study population of high-risk diabetic patients referred for evaluation of CAD differed from the relatively healthy community participants enrolled by Semba et al. Furthermore, the two studies differed in their measurement of PWV, sample size, and AGE type [93]. Several prospective studies have implicated serum levels of multiple AGEs as predictors of development and progression of microvascular complications of DM, including nephropathy and retinopathy [94]. Accordingly, recent efforts have focused on developing drugs specifically directed at blocking steps in AGE formation or collagen cross-linking, but these results have been somewhat disappointing. There have been only a few longitudinal studies examining serial changes in arterial stiffness over time in response to antidiabetic pharmacologic intervention. By far the most studied antidiabetic agents are the peroxisome proliferator activated receptor gamma agonists, thiazolidinediones (TZDs). These studies contained small numbers of patients, the largest involving only 130 subjects. In a meta-analysis, although the pooled estimate demonstrated statistically significant reduction in ankle-brachial PWV with TZDs (either pioglitazone or rosiglitazone), only 1 trial consisting of 30 patients demonstrated independent reduction in arterial stiffness with TZD therapy [95]. Furthermore, clinical outcomes have not improved with these agents, and in fact there is increasing data that TZDs are associated with increased risk in cardiovascular mortality, possibly linked to increasing fluid retention and heart failure [96]. Accordingly, despite some evidence in small clinical trials that TZDs may improve arterial stiffness in diabetic patients, larger trials assessing clinical outcomes are still needed, but may be limited due to cardiovascular toxicity of this class.

377

Finally, while abnormal glucose dysregulation in DM ultimately leads to arterial stiffness, it has also been suggested that arterial stiffness itself may beget abnormal glucose metabolism. The propagation of increased pressure and flow pulsations to small blood vessels that supply target organs such as the brain and kidneys, may over time lead to target organ dysfunction, such as albuminuria and stroke [97,98]. With regard to the pancreas, increase pressure and flow to the pancreatic bed may precipitate pancreatic dysfunction that could lead to DM. Increased penetration of pulsatile energy to skeletal muscle may also result in endothelial dysfunction and metabolic dysregulation, although experimental research in this regard is lacking. Due to the various mechanistic considerations mentioned above, it is possible that increased arterial stiffness may contribute to the development of insulin resistance and DM. Indeed, PP was a predictor of new onset DM in hypertensive patients, and medications that target this hemodynamic effect have been shown to reduce rates of incident DM [99]. Furthermore, in a cohort of patients with untreated essential hypertension, higher PP was independently associated with increasing number of glucose spikes during challenge [100]. Whereas arterial stiffness and elevated PP identify subjects at risk for developing type-2 DM, the possibility that hemodynamic changes related to increased arterial pulsatility play a direct role in the development of DM is intriguing and requires further investigation. 6. Conclusions Accumulating evidence suggests that arterial stiffness is a key component in the pathogenesis of DM, and importantly, that endothelial dysfunction may even occur with early insulin resistance and impaired fasting glucose, before the development of overt DM. Both AGEs and endothelial NO dysregulation play critical roles in the pathogenesis of arterial stiffness. Arterial stiffness is an independent predictor of mortality both in the diabetic population and in the population as a whole, and appears to be linked to the development and progression of target organ dysfunction in patients with DM. While an increased PP independently identifies subjects at risk for the development of DM, whether large artery stiffness and the pulsatile hemodynamic changes that accompany it are involved in the pathogenesis of DM is still unknown. It is still unclear whether improvement in diabetic control is associated with reduction in arterial stiffness, and whether pharmacologic interventions targeting glucose dysregulation or arterial stiffness itself are associated with improved clinical outcomes. Additional prospective studies are needed to further investigate the possible role of large artery stiffness in the pathogenesis of DM, as well as to determine any independent effect of treatment of arterial stiffness on long-term cardiovascular outcomes. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.atherosclerosis.2014.12.023. References [1] C. Vlachopoulos, K. Aznaouridis, C. Stefanadis, Prediction of cardiovascular events and all-cause mortality with arterial stiffness: a systematic review and meta-analysis, J. Am. Coll. Cardiol. 55 (2010) 1318e1327. [2] M.T. Schram, R.M. Henry, R.A. van Dijk, et al., Increased central artery stiffness in impaired glucose metabolism and type 2 diabetes: the Hoorn Study, Hypertension 43 (2004) 176e181. [3] H. Ohnishi, S. Saitoh, S. Takagi, et al., Pulse wave velocity as an indicator of atherosclerosis in impaired fasting glucose: the Tanno and Sobetsu study, Diabetes Care 26 (2003) 437e440. [4] J.Y. Shin, H.R. Lee, D.C. Lee, Increased arterial stiffness in healthy subjects

378

[5] [6]

[7]

[8]

[9] [10]

[11]

[12]

[13]

[14] [15]

[16]

[17]

[18] [19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379 with high-normal glucose levels and in subjects with pre-diabetes, Cardiovasc. Diabetol. 10 (2011) 30. F.S. Fang, M.Y. Liu, X.L. Cheng, et al., Insulin resistance correlates with the arterial stiffness before glucose intolerance, Intern. Med. 53 (2014) 189e194. J.A. Chirinos, P. Segers, T.C. Gillebert, et al., Central pulse pressure and its hemodynamic determinants in middle-aged adults with impaired fasting glucose and diabetes: the Asklepios study, Diabetes Care 36 (2013) 2359e2365. C.H. Li, J.S. Wu, Y.C. Yang, et al., Increased arterial stiffness in subjects with impaired glucose tolerance and newly diagnosed diabetes but not isolated impaired fasting glucose, J. Clin. Endocrinol. Metab. 97 (2012) E658eE662. S. Li, W. Chen, S.R. Srinivasan, et al., Influence of metabolic syndrome on arterial stiffness and its age-related change in young adults: the Bogalusa Heart Study, Atherosclerosis 180 (2005) 349e354. M.E. Safar, F. Thomas, J. Blacher, et al., Metabolic syndrome and age-related progression of aortic stiffness, J. Am. Coll. Cardiol. 47 (2006) 72e75. H. Tomiyama, Y. Hirayama, H. Hashimoto, et al., The effects of changes in the metabolic syndrome detection status on arterial stiffening: a prospective study, Hypertens. Res. Off. J. Jpn. Soc. Hypertens. 29 (2006) 673e678. N. Nakanishi, K. Suzuki, K. Tatara, Clustered features of the metabolic syndrome and the risk for increased aortic pulse wave velocity in middle-aged Japanese men, Angiology 54 (2003) 551e559. P. Pietri, C. Vlachopoulos, G. Vyssoulis, et al., Macro- and microvascular alterations in patients with metabolic syndrome: sugar makes the difference, Hypertens. Res. Off. J. Jpn. Soc. Hypertens. 37 (2014) 452e456. N.K. Sweitzer, M. Shenoy, J.H. Stein, et al., Increases in central aortic impedance precede alterations in arterial stiffness measures in type 1 diabetes, Diabetes Care 30 (2007) 2886e2891. B. Brooks, L. Molyneaux, D.K. Yue, Augmentation of central arterial pressure in type 1 diabetes, Diabetes Care 22 (1999) 1722e1727. I.B. Wilkinson, H. MacCallum, D.F. Rooijmans, et al., Increased augmentation index and systolic stress in type 1 diabetes mellitus, QJM Mon. J. Assoc. Physicians 93 (2000) 441e448. D. Gordin, M. Ronnback, C. Forsblom, et al., Acute hyperglycaemia rapidly increases arterial stiffness in young patients with type 1 diabetes, Diabetologia 50 (2007) 1808e1814. J.D. Cameron, C.J. Bulpitt, E.S. Pinto, et al., The aging of elastic and muscular arteries: a comparison of diabetic and nondiabetic subjects, Diabetes Care 26 (2003) 2133e2138. M.E. Safar, S. Czernichow, J. Blacher, Obesity, arterial stiffness, and cardiovascular risk, J. Am. Soc. Nephrol. 17 (2006) S109eS111. K. Cruickshank, L. Riste, S.G. Anderson, et al., Aortic pulse-wave velocity and its relationship to mortality in diabetes and glucose intolerance: an integrated index of vascular function? Circulation 106 (2002) 2085e2090. E. Kimoto, T. Shoji, K. Shinohara, et al., Regional arterial stiffness in patients with type 2 diabetes and chronic kidney disease, J. Am. Soc. Nephrol. 17 (2006) 2245e2252. H. Taniwaki, T. Kawagishi, M. Emoto, et al., Correlation between the intimamedia thickness of the carotid artery and aortic pulse-wave velocity in patients with type 2 diabetes. Vessel wall properties in type 2 diabetes, Diabetes Care 22 (1999) 1851e1857. R. Ravikumar, R. Deepa, C. Shanthirani, et al., Comparison of carotid intimamedia thickness, arterial stiffness, and brachial artery flow mediated dilatation in diabetic and nondiabetic subjects (The Chennai Urban Population Study [CUPS-9]), Am. J. Cardiol. 90 (2002) 702e707. J.R. Cockcroft, I.B. Wilkinson, M. Evans, et al., Pulse pressure predicts cardiovascular risk in patients with type 2 diabetes mellitus, Am. J. Hypertens. 18 (2005) 1463e1467 discussion 1468e1469. M.T. Schram, N. Chaturvedi, J.H. Fuller, et al., Pulse pressure is associated with age and cardiovascular disease in type 1 diabetes: the Eurodiab Prospective Complications Study, J. Hypertens. 21 (2003) 2035e2044. R.A. van Dijk, J.M. Dekker, G. Nijpels, et al., Brachial artery pulse pressure and common carotid artery diameter: mutually independent associations with mortality in subjects with a recent history of impaired glucose tolerance, Eur. J. Clin. Investig. 31 (2001) 756e763. M.T. Schram, P.J. Kostense, R.A. Van Dijk, et al., Diabetes, pulse pressure and cardiovascular mortality: the Hoorn Study, J. Hypertens. 20 (2002) 1743e1751. J.A. Chirinos, J.G. Kips, M.J. Roman, et al., Ethnic differences in arterial wave reflections and normative equations for augmentation index, Hypertension 57 (2011) 1108e1116. M. Cecelja, P. Chowienczyk, Dissociation of aortic pulse wave velocity with risk factors for cardiovascular disease other than hypertension: a systematic review, Hypertension 54 (2009) 1328e1336. L. De Angelis, S.C. Millasseau, A. Smith, et al., Sex differences in age-related stiffening of the aorta in subjects with type 2 diabetes, Hypertension 44 (2004) 67e71. S.F. Dinneen, H.C. Gerstein, The association of microalbuminuria and mortality in non-insulin-dependent diabetes mellitus. A systematic overview of the literature, Arch. Intern. Med. 157 (1997) 1413e1418. M.E. Molitch, M. Steffes, W. Sun, et al., Development and progression of renal insufficiency with and without albuminuria in adults with type 1 diabetes in the diabetes control and complications trial and the epidemiology of diabetes interventions and complications study, Diabetes Care 33 (2010) 1536e1543.

[32] H. Yokoyama, T. Aoki, M. Imahori, et al., Subclinical atherosclerosis is increased in type 2 diabetic patients with microalbuminuria evaluated by intima-media thickness and pulse wave velocity, Kidney Int. 66 (2004) 448e454. [33] H. Yokoyama, K. Hirasawa, T. Aoki, et al., Brachial-ankle pulse wave velocity measured automatically by oscillometric method is elevated in diabetic patients with incipient nephropathy, Diabet. Med. J. Br. Diabet. Assoc. 20 (2003) 942e945. [34] A. Smith, J. Karalliedde, L. De Angelis, et al., Aortic pulse wave velocity and albuminuria in patients with type 2 diabetes, J. Am. Soc. Nephrol. 16 (2005) 1069e1075. [35] E. Ishimura, H. Taniwaki, T. Tsuchida, et al., Urinary albumin excretion associated with arterial wall stiffness rather than thickness in type 2 diabetic patients, J. Nephrol. 20 (2007) 204e211. [36] S. Aoun, J. Blacher, M.E. Safar, et al., Diabetes mellitus and renal failure: effects on large artery stiffness, J. Hum. Hypertens. 15 (2001) 693e700. [37] R. Bouchi, T. Babazono, M. Mugishima, et al., Arterial stiffness is associated with incident albuminuria and decreased glomerular filtration rate in type 2 diabetic patients, Diabetes Care 34 (2011) 2570e2575. [38] B. Pannier, A.P. Guerin, S.J. Marchais, et al., Stiffness of capacitive and conduit arteries: prognostic significance for end-stage renal disease patients, Hypertension 45 (2005) 592e596. [39] R. Pedrinelli, G. Dell'Omo, G. Penno, et al., Microalbuminuria and pulse pressure in hypertensive and atherosclerotic men, Hypertension 35 (2000) 48e54. [40] A. Oliveras, L. Garcia-Ortiz, J. Segura, et al., Association between urinary albumin excretion and both central and peripheral blood pressure in subjects with insulin resistance, J. Hypertens. 31 (2013) 103e108. [41] S.M. Farasat, C. Valdes, V. Shetty, et al., Is longitudinal pulse pressure a better predictor of 24-hour urinary albumin excretion than other indices of blood pressure? Hypertension 55 (2010) 415e421. [42] A. Fogo, J.A. Breyer, M.C. Smith, et al., Accuracy of the diagnosis of hypertensive nephrosclerosis in African Americans: a report from the African American Study of Kidney Disease (AASK) Trial. AASK Pilot Study Investigators, Kidney Int. 51 (1997) 244e252. [43] A. Juutilainen, S. Lehto, T. Ronnemaa, et al., Retinopathy predicts cardiovascular mortality in type 2 diabetic men and women, Diabetes Care 30 (2007) 292e299. [44] G. Targher, L. Bertolini, L. Zenari, et al., Diabetic retinopathy is associated with an increased incidence of cardiovascular events in type 2 diabetic patients, Diabet. Med. J. Br. Diabet. Assoc. 25 (2008) 45e50. [45] W.J. Kim, C.Y. Park, S.E. Park, et al., The association between regional arterial stiffness and diabetic retinopathy in type 2 diabetes, Atherosclerosis 225 (2012) 237e241. [46] Y.W. Yun, M.H. Shin, Y.H. Lee, et al., Arterial stiffness is associated with diabetic retinopathy in Korean type 2 diabetic patients, J. Prev. Med. Public Health/Yebang Uihakhoe chi 44 (2011) 260e266. [47] O. Ogawa, K. Hiraoka, T. Watanabe, et al., Diabetic retinopathy is associated with pulse wave velocity, not with the augmentation index of pulse waveform, Cardiovasc. Diabetol. 7 (2008) 11. [48] O. Ogawa, C. Hayashi, T. Nakaniwa, et al., Arterial stiffness is associated with diabetic retinopathy in type 2 diabetes, Diabetes Res. Clin. Pract. 68 (2005) 162e166. [49] K. Tanaka, T. Kawai, Y. Saisho, et al., Relationship between stage of diabetic retinopathy and pulse wave velocity in Japanese patients with type 2 diabetes, J. Diabetes Res. 2013 (2013) 193514. [50] M. Rema, V. Mohan, R. Deepa, et al., Association of carotid intima-media thickness and arterial stiffness with diabetic retinopathy: the Chennai Urban Rural Epidemiology Study (CURES-2), Diabetes Care 27 (2004) 1962e1967. [51] D.J. Ewing, I.W. Campbell, B.F. Clarke, The natural history of diabetic autonomic neuropathy, Q. J. Med. 49 (1980) 95e108. [52] H.S. Chen, C.M. Hwu, B.I. Kuo, et al., Abnormal cardiovascular reflex tests are predictors of mortality in type 2 diabetes mellitus, Diabet. Med. J. Br. Diabet. Assoc. 18 (2001) 268e273. [53] E. Zander, B. Schulz, P. Heinke, et al., Importance of cardiovascular autonomic dysfunction in IDDM subjects with diabetic nephropathy, Diabetes Care 12 (1989) 259e264. [54] A.S. Krolewski, J. Barzilay, J.H. Warram, et al., Risk of early-onset proliferative retinopathy in IDDM is closely related to cardiovascular autonomic neuropathy, Diabetes 41 (1992) 430e437. [55] V. Spallone, S. Gambardella, M.R. Maiello, et al., Relationship between autonomic neuropathy, 24-h blood pressure profile, and nephropathy in normotensive IDDM patients, Diabetes Care 17 (1994) 578e584. [56] C. Meyer, F. Milat, B.P. McGrath, et al., Vascular dysfunction and autonomic neuropathy in type 2 diabetes, Diabet. Med. J. Br. Diabet. Assoc. 21 (2004) 746e751. [57] S. Theilade, M. Lajer, F. Persson, et al., Arterial stiffness is associated with cardiovascular, renal, retinal, and autonomic disease in type 1 diabetes, Diabetes Care 36 (2013) 715e721. [58] B.J. Park, H.R. Lee, J.Y. Shim, et al., Association between resting heart rate and arterial stiffness in Korean adults, Arch. Cardiovasc. Dis. 103 (2010) 246e252. [59] B.K. Ha, B.G. Kim, D.H. Kim, et al., Relationships between brachial-ankle pulse wave velocity and peripheral neuropathy in type 2 diabetes, Diabetes Metab. J. 36 (2012) 443e451.

S.B. Prenner, J.A. Chirinos / Atherosclerosis 238 (2015) 370e379 [60] B.C. Suh, P.W. Chung, H.S. Moon, et al., Association between pulse wave velocity and nerve conduction study in diabetic patients, Eur. Neurol. 64 (2010) 219e223. [61] H. Yokoyama, Y. Yokota, J. Tada, et al., Diabetic neuropathy is closely associated with arterial stiffening and thickness in type 2 diabetes, Diabet. Med. J. Br. Diabet. Assoc. 24 (2007) 1329e1335. [62] C.R. Cardoso, M.T. Ferreira, N.C. Leite, et al., Microvascular degenerative complications are associated with increased aortic stiffness in type 2 diabetic patients, Atherosclerosis 205 (2009) 472e476. [63] M. Tanaka, T. Babazono, M. Takeda, et al., Pulse pressure and chronic kidney disease in patients with type 2 diabetes, Hypertens. Res. Off. J. Jpn. Soc. Hypertens. 29 (2006) 345e352. [64] S.T. Knudsen, P.L. Poulsen, K.W. Hansen, et al., Pulse pressure and diurnal blood pressure variation: association with micro- and macrovascular complications in type 2 diabetes, Am. J. Hypertens. 15 (2002) 244e250. [65] E. Laugesen, P. Hoyem, B. Stausbol-Gron, et al., Carotid-femoral pulse wave velocity is associated with cerebral white matter lesions in type 2 diabetes, Diabetes Care 36 (2013) 722e728. [66] O. Ogawa, T. Onuma, S. Kubo, et al., Brachial-ankle pulse wave velocity and symptomatic cerebral infarction in patients with type 2 diabetes: a crosssectional study, Cardiovasc. Diabetol. 2 (2003) 10. [67] S. Mehrabian, M. Raycheva, A. Gateva, et al., Cognitive dysfunction profile and arterial stiffness in type 2 diabetes, J. Neurol. Sci. 322 (2012) 152e156. [68] D.R. Sell, V.M. Monnier, Molecular basis of arterial stiffening: role of glycation e a mini-review, Gerontology 58 (2012) 227e237. [69] R. Meerwaldt, T. Links, C. Zeebregts, et al., The clinical relevance of assessing advanced glycation endproducts accumulation in diabetes, Cardiovasc. Diabetol. 7 (2008) 29. [70] F.J. Tessier, The Maillard reaction in the human body. The main discoveries and factors that affect glycation, Pathol. Biol. 58 (2010) 214e219. [71] A. Prasad, P. Bekker, S. Tsimikas, Advanced glycation end products and diabetic cardiovascular disease, Cardiol. Rev. 20 (2012) 177e183. [72] J.W. Baynes, S.R. Thorpe, Role of oxidative stress in diabetic complications: a new perspective on an old paradigm, Diabetes 48 (1999) 1e9. [73] M. Brownlee, H. Vlassara, A. Kooney, et al., Aminoguanidine prevents diabetes-induced arterial wall protein cross-linking, Science 232 (1986) 1629e1632. [74] International Expert, C, International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes, Diabetes Care 32 (2009) 1327e1334. [75] A.L. Peters, M.B. Davidson, D.L. Schriger, et al., A clinical approach for the diagnosis of diabetes mellitus: an analysis using glycosylated hemoglobin levels. Meta-analysis Research Group on the Diagnosis of Diabetes Using Glycated Hemoglobin Levels, J. Am. Med. Assoc. 276 (1996) 1246e1252. [76] R.J. Koenig, C.M. Peterson, R.L. Jones, et al., Correlation of glucose regulation and hemoglobin AIc in diabetes mellitus, N. Engl. J. Med. 295 (1976) 417e420. [77] J.T. Powell, N. Vine, M. Crossman, On the accumulation of D-aspartate in elastin and other proteins of the ageing aorta, Atherosclerosis 97 (1992) 201e208. [78] S.L. Schnider, R.R. Kohn, Effects of age and diabetes mellitus on the solubility and nonenzymatic glucosylation of human skin collagen, J. Clin. Investig. 67 (1981) 1630e1635. [79] V.M. Monnier, V. Vishwanath, K.E. Frank, et al., Relation between complications of type I diabetes mellitus and collagen-linked fluorescence, N. Engl. J. Med. 314 (1986) 403e408. [80] T.J. Sims, L.M. Rasmussen, H. Oxlund, et al., The role of glycation cross-links in diabetic vascular stiffening, Diabetologia 39 (1996) 946e951. [81] S. Kawashima, The two faces of endothelial nitric oxide synthase in the pathophysiology of atherosclerosis, Endothel. J. Endothel. Cell Res. 11 (2004) 99e107. [82] M.A. Creager, T.F. Luscher, F. Cosentino, et al., Diabetes and vascular disease:

[83]

[84]

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92]

[93]

[94]

[95]

[96]

[97] [98]

[99]

[100]

379

pathophysiology, clinical consequences, and medical therapy: part I, Circulation 108 (2003) 1527e1532. X. Du, D. Edelstein, S. Obici, et al., Insulin resistance reduces arterial prostacyclin synthase and eNOS activities by increasing endothelial fatty acid oxidation, J. Clin. Investig. 116 (2006) 1071e1080. D.G. Brillante, A.J. O'Sullivan, M.T. Johnstone, et al., Arterial stiffness and haemodynamic response to vasoactive medication in subjects with insulinresistance syndrome, Clin. Sci. 114 (2008) 139e147. S. Vehkavaara, A. Seppala-Lindroos, J. Westerbacka, et al., In vivo endothelial dysfunction characterizes patients with impaired fasting glucose, Diabetes Care 22 (1999) 2055e2060. A. Avogaro, F. Piarulli, A. Valerio, et al., Forearm nitric oxide balance, vascular relaxation, and glucose metabolism in NIDDM patients, Diabetes 46 (1997) 1040e1046. D.G. Brillante, A.J. O'Sullivan, L.G. Howes, Arterial stiffness in insulin resistance: the role of nitric oxide and angiotensin II receptors, Vasc. Health Risk Manag. 5 (2009) 73e78. B.K. Kilhovd, A. Juutilainen, S. Lehto, et al., Increased serum levels of advanced glycation endproducts predict total, cardiovascular and coronary mortality in women with type 2 diabetes: a population-based 18 year follow-up study, Diabetologia 50 (2007) 1409e1417. J.W. Nin, A. Jorsal, I. Ferreira, et al., Higher plasma levels of advanced glycation end products are associated with incident cardiovascular disease and all-cause mortality in type 1 diabetes: a 12-year follow-up study, Diabetes Care 34 (2011) 442e447. J.W. Nin, A. Jorsal, I. Ferreira, et al., Higher plasma soluble Receptor for Advanced Glycation End Products (sRAGE) levels are associated with incident cardiovascular disease and all-cause mortality in type 1 diabetes: a 12year follow-up study, Diabetes 59 (2010) 2027e2032. M.T. Schram, C.G. Schalkwijk, A.H. Bootsma, et al., Advanced glycation end products are associated with pulse pressure in type 1 diabetes: the EURODIAB Prospective Complications Study, Hypertension 46 (2005) 232e237. R.D. Semba, S.S. Najjar, K. Sun, et al., Serum carboxymethyl-lysine, an advanced glycation end product, is associated with increased aortic pulse wave velocity in adults, Am. J. Hypertens. 22 (2009) 74e79. K.B. Won, H.J. Chang, S.H. Park, et al., High serum advanced glycation endproducts predict coronary artery disease irrespective of arterial stiffness in diabetic patients, Korean Circ. J. 42 (2012) 335e340. V.M. Monnier, D.R. Sell, S. Genuth, Glycation products as markers and predictors of the progression of diabetic complications, Ann. N. Y. Acad. Sci. 1043 (2005) 567e581. D.R. Webb, M.J. Davies, L.J. Gray, et al., Searching for the right outcome? A systematic review and meta-analysis of controlled trials using carotid intima-media thickness or pulse wave velocity to infer antiatherogenic properties of thiazolidinediones, Diabetes Obes. Metab. 12 (2010) 124e132. R.W. Nesto, D. Bell, R.O. Bonow, et al., Thiazolidinedione use, fluid retention, and congestive heart failure: a consensus statement from the American Heart Association and American Diabetes Association, Circulation 2003 (108) (October 7, 2003) 2941e2948. M.F. O'Rourke, J. Hashimoto, Mechanical factors in arterial aging: a clinical perspective, J. Am. Coll. Cardiol. 50 (2007) 1e13. M.L. Ford, L.A. Tomlinson, T.P. Chapman, et al., Aortic stiffness is independently associated with rate of renal function decline in chronic kidney disease stages 3 and 4, Hypertension 55 (2010) 1110e1115. S. Yasuno, K. Ueshima, K. Oba, et al., Is pulse pressure a predictor of newonset diabetes in high-risk hypertensive patients?: A subanalysis of the Candesartan Antihypertensive Survival Evaluation in Japan (CASE-J) trial, Diabetes Care 33 (2010) 1122e1127. F. Anan, T. Masaki, T. Eto, et al., Postchallenge plasma glucose and glycemic spikes are associated with pulse pressure in patients with impaired glucose tolerance and essential hypertension, Hypertens. Res. Off. J. Jpn. Soc. Hypertens. 31 (2008) 1565e1571.