Risk Factors for Metabolic Syndrome Independently Predict Arterial Stiffness and Endothelial Dysfunction in Patients With Chronic Kidney Disease and Minimal Comorbidity

The rationale and study design have been reported in detail elsewhere (4). In brief, subjects were recruited from the renal outpatient clinic at the Royal Infirmary of Edinburgh. They were categorized into the five stages of CKD on the basis of the Kidney Disease Outcome Quality Initiative (K/DOQI) classification (20). Age-matched healthy volunteers were recruited from the community as a control group.

The inclusion criteria were as follows: male or female CKD patients, 18–65 years old, and clinic BP ≤160/100 mmHg, whether or not on antihypertensive medication. We excluded patients with a renal transplant or on dialysis, systemic vasculitis or connective tissue disease, a history of established CVD, peripheral vascular disease, diabetes, respiratory or neurologic disease, and current alcohol abuse or pregnancy and those treated with an organic nitrate or β-agonist. Patients continued their usual medications until the study morning. Smokers and hypercholesterolemic patients were not excluded, but the latter had to be established on statin medication, with good cholesterol control, for at least 3 months before taking part in the study.

MS was diagnosed according to the criteria from the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel [ATP] III) (10,21). The diagnosis of MS was made when subjects had three or more risk factors for MS. Subjects with zero to one and two risk factors for MS were classified as no MS and risk for developing MS, respectively (Supplementary Table 1).

Brachial systolic and diastolic BP (SBP and DBP) were recorded in duplicate, with an appropriately sized cuff, using a validated oscillometric sphygmomanometer (Omron HEM-705CP) (22), and values were presented as an average of two recordings. BMI was calculated as weight (kg)/(height [m])². Waist and hip circumferences (cm) were measured on a subject in a standing position with feet 15 cm apart. Waist circumference was measured at the midpoint between the iliac crest and lowest rib. Hip circumference was measured at the midpoint between the waist and groin. During the measurement, the tape measure was parallel to the ground.

Arterial stiffness was assessed by measuring carotid-femoral PWV (CF-PWV), as previously described (23), using the SphygmoCor apparatus (SphygmoCor BPAS-1; AtCor Medical, Sydney, Australia). Central augmentation index (cAIx) was estimated using the same system (23). Brachial artery flow-mediated dilation (FMD) was used to assess endothelium-dependent vasomotor function as described elsewhere (24). FMD was quantified as a percentage change from baseline in brachial artery diameter after 5 min of forearm ischemia. Endothelium-independent vasomotor function was assessed using 25 μg nitroglycerine (NTG) by sublingual administration (25).

Creatinine clearance, as an estimate of glomerular filtration rate (eGFR), was calculated according to the Cockcroft and Gault (C&G) equation (26): (140 − age [years] × weight [kg] × 1.23 for male or 1.05 for female)/serum creatinine (μmol/L). The C&G equation was selected to assess renal function in this study because it is more accurate than the Modification of Diet in Renal Disease (MDRD) equation when used to assess mild renal insufficiency (27). It was further corrected by body surface area.

Data were statistically analyzed using SPSS program for Windows (SPSS 15.0; SPSS Inc., Chicago, IL). Descriptive data are given as mean ± SD unless otherwise stated. Means of the categorical data (subjects without MS, subjects with risk of developing MS, and subjects with MS) were compared by one-way ANOVA. Continuous data (risk factors for MS as zero to five risk factors) were analyzed by correlation coefficients calculated using the Pearson method. Stepwise linear regression was used for multivariate analysis. A significant level was taken as P value <0.05.

CKD patients (n = 113) and age-matched non-CKD control subjects (n = 23) were enrolled into the study. Baseline characteristics of the studied subjects are given in Supplementary Table 2. Causes of CKD and medication used by the patients are described in Supplementary Table 3.

Subjects were classified into three categories according to the number of risk factors for MS (see 1research design and methods). Twenty-six subjects (19%) had MS and 27 (20%) were defined as at risk for developing MS. All three categories were comparable in respect to age and eGFR. As expected, subjects with MS had a higher BMI, waist circumference, SBP, DBP, plasma glucose, and triglycerides and lower HDL cholesterol compared with those without MS or those at risk for developing it (Table 1). With regard to the relationship of the risk factors for MS to renal function, only SBP increased as eGFR declined (r² = 0.11; P < 0.01); waist circumference, DBP, plasma glucose, triglycerides, and HDL cholesterol showed no relationship to renal function. CF-PWV was higher in the MS group (Table 1 and Fig. 1A and C) whereas cAIx showed no relationship to MS (Table 1). FMD was lower in the MS group and subjects at risk for MS compared with those without MS (Table 1 and Fig. 1B and D) but this did not reach statistical significance. The endothelium-independent response to NTG had no relationship to MS (Table 1).

Univariate analysis to assess the relationship of CF-PWV to individual MS risk factors, the number of MS risk factors, and related parameters that are not included in the National Cholesterol Education Program (NCEP) ATP III criteria (eGFR, age, BMI, and waist-to-hip ratio) was performed. CF-PWV only correlated with eGFR when all subjects were considered together (r² = 0.07; P < 0.01) but not when split into the three groups according to the presence, risk, or absence of MS (Fig. 1A). However, PWV in subjects with MS was significantly higher than in subjects without MS (Fig. 1A and C). With regard to the individual risk factors for MS, CF-PWV increased with waist circumference, waist-to-hip ratio, SBP, and plasma glucose (Fig. 2A–D). Additionally, CF-PWV correlated with age (r² = 0.25; P < 0.01), BMI (r² = 0.06; P < 0.01), DBP (r² = 0.07; P < 0.01), and plasma triglycerides (r² = 0.05; P < 0.05) but not sex, smoking status, or HDL cholesterol.

In multivariate analysis, when renal function and conventional risk factors were entered into the model, age and sex were independent predictors of arterial stiffness (Table 2, model 1). When either the presence of MS or the number of risk factors for MS that a subject had were considered, both of them, together with age and sex, independently predicted CF-PWV (Table 2, models 2 and 3). When the individual risk factors for MS were considered, sex was replaced by waist circumference, SBP, and DBP as independent predictors of CF-PWV (Table 2, model 4).

In univariate analysis, FMD correlated with eGFR when all subjects were considered together (r² = 0.04; P < 0.05) but not when subjects were divided into the three MS groups. FMD did not change according to the presence, risk, or absence of MS (Fig. 1B and D). FMD correlated with waist circumference, waist-to-hip ratio, and SBP but not plasma glucose (Fig. 2E–H). Additionally FMD did not correlate with age, BMI, DBP, plasma triglycerides, HDL cholesterol, or smoking status.

In multivariate analysis assessing renal function and conventional risk factors, eGFR was an independent predictor of FMD (Table 2, model 1). Although the presence or absence of MS did not predict endothelial dysfunction, the number of MS risk factors did (Table 2, models 2 and 3). When the risk factors for MS were considered individually in the model, SBP and DBP alone were independent predictors of FMD (Table 2, model 4).

In a cohort of subjects with minimal comorbidity and an eGFR ranging from 8 to 154 mL/min/1.73 m², we have previously shown that renal function is related to an increase in arterial stiffness and endothelial dysfunction, as measured by CF-PWV and FMD, respectively (4). We now show that subjects in this cohort with MS have increased arterial stiffness compared with those without MS. Importantly, this is independent of the level of renal function (eGFR). Both the presence of MS and the number of MS risk factors are independent predictors of arterial stiffness in this cohort. Furthermore, when risk factors for MS were considered individually, waist circumference and blood pressure were determinants of arterial stiffness, independent of renal function, age, sex, and smoking status.

Our study cohort was carefully selected to have low comorbidity. Thus, the prevalence of MS in this CKD population is lower than previously reported in dialysis (19,28), hypertensive (15), and diabetic patients (16) but is comparable to healthy subjects (18 vs. 10–19%) (14). Interestingly, despite this selection bias against MS and its risk factors, we have shown that, irrespective of renal function, there is an increase in arterial stiffness in our subjects with MS or risk factors for MS, and that these independently predict arterial stiffness when considered alongside conventional risk factors. This finding is in keeping with previous data from healthy subjects (13,14,29,30), hypertensive patients (15), and patients with diabetes (16).

Our data also show that waist circumference is a predictor of arterial stiffness in this population and this is independent of renal function, age, BP, and sex. Our analysis considered the whole cohort of 113 CKD patients and 23 age-matched, non-CKD control subjects, but waist circumference remains an independent predictor of arterial stiffness when the CKD patients are considered alone. This has previously only been shown in hypertensive patients (15). Obesity is associated both with increased arterial stiffness (31) and an increased risk of CVD (32). Our data suggest that, in patients with CKD, waist circumference, a marker of central obesity, may be a better surrogate for arterial stiffness than BMI or cholesterol subtypes such as HDL cholesterol and triglycerides.

In the current study, subjects with MS did not have significantly impaired endothelial function compared with those without MS, regardless of renal function. This is similar to data from studies in healthy subjects (33,34), those at risk for developing diabetes (35,36), and those with peripheral vascular disease (37). However, in a study of ∼1,000 elderly subjects, including those with CVD and diabetes, where endothelial function was assessed using both an invasive forearm technique and FMD (38), the former showed significantly impaired endothelial function in the MS group. The invasive forearm technique is considered to assess vascular function of the resistance arteries, whereas FMD is a measure of conduit artery function. Thus, it is possible that MS is associated with a predominant dysfunction of the resistance vessels and, hence, by using FMD, we did not detect significant endothelial dysfunction in our study cohort. We also observed no association between the number of risk factors for MS and FMD. This is in contrast to a previous study in subjects at risk for developing diabetes (36), but this may in part be explained by genetic influences in diabetes (and insulin resistance states) not currently considered to be of importance in the majority of CKD patients.

Of all the risk factors studied here, conventional and MS related, only BP is an independent predictor of endothelial dysfunction (Table 2, model 4). This result both confirms and contradicts those of previous studies. Similar to our findings, Scuteri et al. (35) found BP (both SBP and DBP) to be an independent predictor of endothelial dysfunction in normoglycemic first-degree relatives of patients with diabetes. However, Kovaite et al. (29) found this not to be the case when studying 186 “asymptomatic subjects without overt cardiovascular disease.” Notably, one-third of subjects studied fulfilled criteria for MS. A study in CKD has shown that, similar to the current data, renal function is a predictor of endothelial dysfunction (39). However, this study was performed in CKD patients with diabetes and this may act as a significant confounder. Furthermore, in that study, an invasive forearm technique measured endothelial function in a different vascular bed compared with the technique of FMD used in the current study.

Interestingly, we found an inverse association between DBP and CF-PWV (Table 2) and a positive association between DBP and FMD (Table 2). However, these may be explained by an effect of pulse pressure, which is positively associated with CF-PWV and inversely related to FMD, since both SBP and DBP were entered into the analysis.

We recognize some limitations to the current study. There are several criteria for the diagnosis of MS (14). As our patients were not diabetic, we cannot use the criteria proposed by the World Health Organization and, therefore, the evaluation of an effect of insulin resistance to arterial stiffness and endothelial dysfunction is limited and does not allow data comparison with other studies using World Health Organization criteria. Also, an increased girth has the potential to artifactually increase the measurement entered for the distance traveled by the pulse wave (compared with the actual distance) in calculating PWV. Thus, the reported PWV will be higher than the “true” value. It is possible that this is responsible, in part, for the higher PWV in the MS group in this study. However, our subjects were chosen because of low comorbidity and are not typical of “general” CKD patients (including patients with diabetes) particularly in their body habitus. Median BMI was 26.9 kg/m², and the patient with the highest BMI did not have MS by ATP III criteria. Additionally, the 54 patients whose waist measurement was above the defining level for a risk factor for MS were spread throughout the three groups: 16 patients were in the “no MS” group, 17 in the “risk of MS” group, and 21 in the “MS” group. Some of the medications taken by our patients, such as angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, β-blockers, and statins, may have had effects on both CF-PWV and FMD. However, all patients were stabilized on their therapies and did not receive their medications on the study day until after measurements were made. The low number of smokers in this study may be responsible for the lack of expected relationships between smoking status and both arterial stiffness and endothelial dysfunction; although when only the CKD patients are considered in the multivariate analysis, smoking is an independent predictor of PWV in models 2 and 3. Finally, as this was a cross-sectional study, we cannot fully define the causal relationships behind the associations described, though they do have a sound mechanistic basis.

In summary, the current study shows that, in the absence of CVD and diabetes and irrespective of renal function, age, smoking status, and sex, CKD patients with MS have increased arterial stiffness compared with those without MS but no difference in endothelial function. Although BP remains the strongest determinant of both arterial stiffness and endothelial dysfunction, the presence of MS or its risk factors is also an important determinant of arterial stiffness in this CKD cohort. Importantly, waist circumference is an independent predictor of arterial stiffness in this cohort; hence, education on lifestyle changes and weight management interventions offer a different therapeutic perspective, complementary to drug management of traditional cardiovascular risk factors, such as hypertension and hypercholesterolemia, that should not be overlooked in patient care. Given these currently noncomorbid patients in this cohort only had modest increases in BMI, the early and aggressive control of weight should be a focus for intervention. As this is not related to renal function, it provides an additional target for therapies aimed at improving cardiovascular outcomes at all stages of CKD.

This study was funded by a British Heart Foundation project grant (PG/05/91). P.L. was supported by the Royal Thai Government for her PhD training. This work was undertaken within Edinburgh’s British Heart Foundation Centre of Research Excellence.

No potential conflicts of interest relevant to this article were reported.

P.L. designed and performed the study and prepared the manuscript. N.D. designed the study and revised and edited the manuscript. V.M. and D.K. performed the study. D.J.W. and J.G. designed and supervised the study and revised and edited the manuscript. J.G. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

This study was presented in poster form at the 10th International Conference on Endothelin, Bergamo, Italy, 16–19 September 2007, and the American Society of Nephrology Annual Meeting, San Francisco, California, 31 October–5 November 2007.