Article Text

Original article
Systemic inflammation and brachial artery endothelial function in the Multi-Ethnic Study of Atherosclerosis (MESA)
1. Shepard D Weiner1,
2. Hanna N Ahmed1,
3. Zhezhen Jin2,
4. Mary Cushman3,
5. David M Herrington4,
6. Jennifer Clark Nelson5,
7. Marco R Di Tullio1,
8. Shunichi Homma1
1. 1Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York, USA
2. 2Department of Biostatistics, Joseph P. Mailman School of Public Health, Columbia University, New York, New York, USA
3. 3Departments of Medicine and Pathology, University of Vermont College of Medicine, Burlington, Vermont, USA
4. 4Department of Medicine, Wake Forest University School of Medicine, Winston-Salem, North Carolina, USA
5. 5Biostatistics Unit, Group Health Research Institute and Department of Biostatistics, University of Washington, Seattle, Washington, USA
1. Correspondence to Dr Shepard D Weiner, Division of Cardiology, Columbia University College of Physicians and Surgeons, Columbia University Medical Center, New York Presbyterian Hospital, Milstein Family Heart Center, 4th Floor, 173 Fort Washington Avenue, New York, NY 10032, USA; sw2150{at}columbia.edu

## Abstract

Background and objective Inflammation and endothelial dysfunction have been implicated in the pathogenesis of atherosclerotic vascular disease. Brachial artery flow-mediated dilation (FMD) is a reliable, non-invasive method of assessing endothelial function. We hypothesised that increased levels of systemic inflammatory markers are associated with impaired endothelial function as assessed by FMD in a multi-ethnic cohort.

Methods We assessed brachial artery FMD in 3501 participants (1739 men, 1762 women; median age 61 years) in the Multi-Ethnic Study of Atherosclerosis and measured serum concentrations of interleukin (IL)-6, C reactive protein (CRP) and tumour necrosis factor (TNF)-α receptor 1. Spearman correlation coefficients were used to evaluate the association of each inflammatory marker with FMD, adjusting for the effect of other variables associated with FMD.

Results There was a significant inverse correlation between IL-6 levels and FMD (−0.042; p=0.02) after adjustment for age, gender, race/ethnicity, education, income, low-density lipoprotein, diabetes, glucose, hypertension status and treatment, waist circumference, triglycerides, baseline brachial diameter, recent infection and use of medications that may alter inflammation. There was no significant correlation between CRP and FMD (0.008; p=0.64) or TNF-α receptor 1 and FMD (0.014; p=0.57). There was no evidence of effect modification by race/ethnicity.

Conclusions In this multi-ethnic cohort, increased levels of the pro-inflammatory cytokine IL-6 were associated with impaired endothelial function assessed by FMD. Elevated IL-6 levels may reflect a state that promotes vascular inflammation and development of subclinical atherosclerosis independent of traditional cardiovascular risk factors.

• Endothelium
View Full Text

## Introduction

Advancements in understanding of the pathobiology of atherothrombosis have implicated inflammation as a central contributor to atherosclerotic vascular disease. Inflammation plays a major role in all stages of atherogenesis and its systemic complications.1 In addition to its role in established cardiovascular disease, inflammation contributes to the earliest steps in atherogenesis, supporting the concept of using pro-inflammatory biomarkers to predict future cardiovascular risk.2

Ultrasound assessment of brachial artery flow-mediated dilation (FMD) during reactive hyperaemia is a reproducible method for the assessment of endothelial dysfunction.3 Endothelial dysfunction is a frequent precursor of early atherogenesis.4 Impaired endothelium-dependent dilation in the coronary circulation is associated with coronary risk factors and atherosclerosis and improves with risk reduction therapies.5 Impaired brachial artery FMD is predictive of endothelial dysfunction in the coronary circulation.6 Therefore, brachial artery FMD has emerged as an attractive non-invasive technique for the study of endothelial function as a marker of subclinical atherosclerosis.

There is prior evidence to suggest that IL-6 may play a role in the pathobiology of atherogenesis. IL-6 gene transcripts have been found to be expressed in human atherosclerotic lesions.7 The main biological effect of IL-6 is regulation of the acute-phase response. During acute inflammation, IL-6 promotes activation of T cells and differentiation of B cells, and regulates liver production of acute phase reactants such as CRP. Additionally, IL-6 stimulates platelet aggregation and expression of tissue factor.8 Elevated levels of interleukin-6 (IL-6) have been shown to predict the development of type II diabetes9 and are associated with increased risk of future cardiovascular events after adjustment for traditional cardiovascular risk factors in healthy individuals.10

Previous studies examining the relationship between inflammatory biomarkers and endothelial dysfunction have shown conflicting results.11–14 To date, however, no large multi-ethnic study has examined the relationship between inflammatory biomarkers and endothelial dysfunction as assessed by FMD. We thus sought to investigate whether IL-6 or CRP levels were associated with vascular endothelial function in a large multi-ethnic community-based sample while adjusting for known cardiovascular risk factors.

Tumour necrosis factor (TNF)-α is a pro-inflammatory cytokine that also induces secretion of CRP from the liver. While CRP has been well studied as an independent predictor of cardiovascular events, few studies have examined the relationship between TNF-α and endothelial function or cardiovascular events, and those that have show mixed results.15–17 To extend prior findings, we also studied circulating TNF-α receptor levels. TNF-α receptor levels are more stable and can be measured with greater sensitivity and reliability than TNF-α itself.18

## Methods

### Study cohort

The Multi-Ethnic Study of Atherosclerosis (MESA) was initiated in July 2000 to investigate the prevalence, correlates and progression of subclinical atherosclerosis in a population-based sample of individuals without known cardiovascular disease.19 This is a population-based study of 6814 women and men aged 45–84 years old recruited from six US communities (Baltimore, Maryland; Chicago, Illinois; Forsyth County, North Carolina; Los Angeles, California; New York, New York; and St. Paul, Minnesota). Cohort participants were 38% Caucasian (n=2624), 28% African American (n=1895), 22% Hispanic (n=1492) and 12% Chinese (n=803). Details of the MESA recruitment strategy have previously been reported.20 The MESA study was approved by institutional review boards at each site, and all participants gave written informed consent.

Medical history, anthropometric measurements and laboratory data were taken from the first examination of the MESA cohort (July 2000 to August 2002). FMD determination and blood collection for inflammatory marker assessment were performed on the same day. Information about age, gender and ethnicity were obtained by questionnaires. Education was classified into three groups: less than high school, complete high school or equivalent certification, and complete college or more. Annual household income was bracketed into three categories: <$20 000,$20 000–$49 999, and >$50 000. Current smoking was defined as having smoked a cigarette in the last 30 days. Diabetes was defined as having a fasting glucose of ≥126 mg/dL, use of hypoglycaemic medications or self-report of diabetes. Resting blood pressure was measured three times in the seated position, and the average of the second and third readings was recorded. Hypertension status was defined as a systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg. Waist circumference was measured in centimetres. Triglycerides and high-density lipoprotein (HDL) cholesterol were measured from blood samples obtained after a 12 h fast. Low-density lipoprotein (LDL) cholesterol was estimated by the Friedewald equation. Participants were categorised as having a recent infection if they reported cold/flu, sinus, urinary or tooth infection, bronchitis or pneumonia in the preceding 2 weeks. Medication use that alters inflammation was defined as current use of lipid-lowering medications, non-steroidal anti-inflammatory drugs or aspirin.

Brachial artery FMD was documented with the use of ultrasound in the MESA study during the first examination. Of the 6814 MESA participants, 6489 (95.2%) had successful FMD testing. Participants were excluded from the FMD examination if they had uncontrolled hypertension (systolic blood pressure >180 mm Hg or diastolic blood pressure >110 mm Hg) or blood pressures in the left and right arms that differed by >15 mm Hg (n=158), a history of Raynaud phenomenon (n=55), a congenital abnormality of the arm or hand (n=12) or mastectomy (n=100). Brachial ultrasound videotapes were analysed at the Wake Forest University Cardiology Image Processing Laboratory. Images from 5731 participants were of sufficient quality for reading. Due to financial constraints, images were read for a subset of these participants (n=3501). These participants were a nested case-cohort that included all participants who had an adjudicated cardiovascular event (n=188), and a random sample of the remaining participants in the MESA study (n=3313). Of these 3501 participants, 1739 were men and 1762 were women.

### Measurement of inflammatory biomarkers

Inflammatory markers examined include IL-6, CRP and TNF-α receptor 1 (R1). Standardised methods were used for blood collection, processing and shipping, with fasting morning phlebotomy and use of a central laboratory (Laboratory for Clinical Biochemistry Research, University of Vermont, Burlington).21 IL-6 and TNF-α R1 were measured by ultrasensitive ELISA from R&D Systems (Minneapolis, Minnesota, USA). TNF-α R1 was successfully measured in 2871 participants of the Exam 1 MESA cohort. These participants correspond to 2880 ‘group 4’ participants who were selected for an early ancillary candidate gene effort, where participants were selected equally from the four race/ethnic groups (n=720 selected at random from each race/ethnic group). High-sensitivity CRP was measured by nephelometry using the BNII nephelometer (N High Sensitivity CRP, Dade Behring, Deerfield, Illinois, USA).

### Determination of flow-mediated dilation

Participants were examined in the supine position after 15 min of rest, after at least 6 h fasting and at least 6 h without smoking. An automated sphygmomanometer (Dinamap device) was used to monitor blood pressure and pulse in the left arm at 5 min intervals throughout the examination. A standard blood pressure cuff was positioned around the right arm, two inches below the antecubital fossa, and the artery was imaged 5–9 cm above the antecubital fossa. A linear-array multifrequency transducer operating at 9 MHz (GE Logiq 700 Device) was used to acquire images of the right brachial artery. After baseline images were obtained, the cuff was inflated to 50 mm Hg above the participant's systolic blood pressure for 5 min. Digitised images of the right brachial artery were captured continuously for 30 s before cuff inflation and for 2 min beginning immediately before cuff deflation to document the vasodilator response. Semiautomated readings (media-adventitial interfaces to media-adventitial interfaces) of these digitised images generated the baseline and maximum diameters of the brachial artery from which % FMD was determined as follows: % FMD=((maximum diameter − baseline diameter)/baseline diameter)×100.

To evaluate intra-reader reproducibility for baseline diameter, maximum diameter and % FMD, ultrasound examinations for 40 MESA participants (32 men and 8 women; 18 Caucasian, 2 Chinese, 10 African American and 10 Hispanic) were reanalysed. The intraclass correlation coefficients were 0.99, 0.99 and 0.93, for Caucasian, African American and Hispanic groups, respectively. The intraclass correlation coefficient was not calculated for the Chinese group due to small sample size. A more detailed description of the scanning and reading protocol can be found at the MESA website (http://www.mesa-nhlbi.org).

### Statistical analysis

Either medians and ranges or mean and SD were calculated for continuous variables and proportions for categorical variables. Either the two-sided Wilcoxon rank sum test or the two-sided t test was used for the comparison for continuous variables, and the χ2 test was used for the comparison for categorical variables. Since the distributions of FMD and inflammatory markers were skewed, we used Spearman correlation coefficients to evaluate the association of each inflammatory marker with FMD. We estimated raw Spearman correlations, Spearman partial correlations that accounted for age and gender, and Spearman partial correlations that accounted for age, gender, race/ethnicity, education, income, LDL, diabetes, glucose, hypertension status and treatment, waist circumference, triglycerides, baseline brachial diameter, recent infection and use of medications. These included clinical covariates previously reported to correlate with FMD (cigarette use, hypertension status, fasting glucose, diabetes mellitus, LDL cholesterol, HDL cholesterol, triglycerides and waist circumference). Recent infection and use of medications were also included as they were known to alter inflammation. Statistical significance was determined at an α-level of 0.05 with two-sided tests. Statistical analyses were conducted with SAS V.9.3.

## Results

### Participant characteristics, measures of vascular function and biomarkers

Clinical characteristics of the 3501 participants with FMD measurements are provided in table 1. IL-6 assays were available for 97% of these participants (n=3404), CRP for 99% (n=3481) and TNF-α R1 for 52% (n=1812).

Table 1

Clinical characteristics of the study sample

The median (range) baseline brachial artery diameter was 3.8 mm (2.2–6.0 mm) for women and 4.8 mm (2.5–7.1 mm) for men. The median (range) FMD for women was 0.16 mm (0–0.74 mm) or 4.2% (0–24.8%). The median (range) FMD for men was 0.17 mm (−0.07 to 0.63 mm) or 3.5% (0–14.8%). The summary statistics for these measures of vascular function and for inflammatory markers are displayed in table 2. In adjusted Spearman correlation analysis, the measured variables that had the strongest correlation with FMD were baseline brachial artery diameter (−0.379; p<0.0001), age (−0.242; p<0.0001) and waist circumference (0.083; p<0.0001). Unadjusted Spearman correlation coefficients among the three markers of inflammation were positively correlated (IL-6 and CRP 0.536, p<0.0001; IL-6 and TNF-α R1 0.336, p<0.0001; CRP and TNF-α R1 0.237, p<0.0001). For further descriptive information, see the online supplementary table.

Table 2

Markers of inflammation and measures of vascular function

### Inflammatory markers and vascular function

Table 3 shows the correlations between inflammatory markers and brachial artery FMD. In univariate analysis, there was a significant inverse correlation between IL-6 levels and FMD (−0.115; p<0.001) and between TNF-α R1 and FMD (−0.096; p<0.001), but no correlation between CRP and FMD (−0.010; p=0.58). In multivariate analysis, the inverse relationship between elevated IL-6 levels and brachial artery FMD remained significant (−0.042; p=0.02), but there was no correlation between TNF-α R1 and FMD (0.014; p=0.570) or CRP and FMD (0.008; p=0.64). The multivariable modelling did account for some of the relationship between IL-6 levels and FMD.

Table 3

Relationship between flow-mediated dilatation and markers of inflammation

### Inflammatory biomarkers and FMD by race/ethnic group

The correlation between IL-6 and FMD was significant among Caucasian, Chinese and Hispanic race/ethnic groups in univariate analyses. In complete multivariate analysis however, there was no evidence for a difference in the association of IL-6 with FMD by race/ethnicity; that is, there was no evidence for effect modification by race. Although it did not reach statistical significance, the association was larger among non-Caucasian groups, particularly for Hispanics (see table 4). For African Americans, the correlation between TNF-α R1 and FMD was significant in the complete multivariate model. There was no significant association of CRP with FMD in any race/ethnic group.

Table 4

Relationship between flow-mediated dilation and inflammatory markers by race/ethnic group

## Discussion

In this large multi-ethnic community-based sample of adults 45–84 years of age who were free of clinical cardiovascular disease, higher levels of the pro-inflammatory cytokine IL-6 were associated with endothelial dysfunction as assessed by impaired FMD on multivariate analyses that adjusted for traditional cardiovascular risk factors. The other inflammatory biomarkers, CRP and TNF-α R1, were not related to FMD on multivariable analysis. Although the association between elevated IL-6 levels and impaired FMD remained present on multivariable analysis, part of the association was mediated by traditional cardiovascular risk factors as well as baseline brachial artery diameter.

Although the association between elevated IL-6 levels and impaired FMD in our study was not individually significant in any of the race/ethnic groups studied, the association was weakest in Caucasians. This trend, along with other data that demonstrate ethnic variations in atherosclerosis and IL-6,22 seems to suggest that the significant association between elevated IL-6 levels and impaired FMD in this study might at least partially relate to the inclusion of non-Caucasian race/ethnic groups in the MESA cohort. While our study did not show a significant relationship between TNF-α R1 and FMD overall, in the subgroup of African Americans, TNF-α R1 was significantly associated with FMD in both the univariate and the complete multivariable models. Whether this occurred by chance or is a true association is not known. This is a novel finding which warrants further study.

Our study results are in line with recent publications on inflammatory biomarkers and endothelial function in community-based cohorts, which have demonstrated weak or no associations of circulating inflammatory biomarkers with FMD. Two previous community-based cohort studies involving the largely Caucasian Framingham Offspring Study have examined the relationship between inflammatory biomarkers and endothelial dysfunction. One study observed modest unadjusted correlations between FMD and the inflammatory markers IL-6 and CRP that were rendered non-significant after adjustment for cardiovascular disease risk factors in 2701 participants. In this study, CRP and IL-6 were inversely related to FMD and reactive hyperaemia in the unadjusted models. After multivariable adjustment, CRP and IL-6 were no longer correlated with FMD, but were still correlated with reactive hyperaemia though the association was weaker.12 Another study conducted in the same cohort a few years later confirmed the finding of no association of CRP with FMD.23 The Framingham investigators did not evaluate TNF-α R1. The lack of association between CRP and brachial artery FMD in the current study is consistent with the results of those studies. However, we found a significant association between IL-6 and FMD that was not observed in the first Framingham Offspring study. In another study of healthy Caucasian men, the investigators also found a significant negative independent association of IL-6 with endothelium-dependent vasodilation in their multivariable analysis. In this study, IL-6 was also found to be negatively associated with insulin sensitivity.11

If further studies confirm these findings, therapies that target IL-6, either indirectly or directly, may be of use. A significant reduction in IL-6 levels and improvement in endothelium-dependent vasodilation of the brachial artery has been demonstrated in patients treated with atorvastatin,24 supporting the anti-inflammatory role of statin pleiotropy. There is also evidence that the role IL-6 plays in promoting endothelial dysfunction is related to the renin–angiotensin system. Treatment with the angiotensin receptor blocker (ARB) irbesartan for 4 weeks resulted in statistically significant reductions in plasma levels of IL-6 and a 67% increase in FMD of the brachial artery compared to subjects who received placebo.25 Another study demonstrated that in post-myocardial infarction subjects with impaired left ventricular systolic function, baseline FMD was low and levels of TNF-α and CRP were high. After 8 and 12 weeks, in the group treated with quinapril, but not in the group treated with enalapril, FMD significantly improved and levels of TNF-α and CRP significantly decreased.26

In a cross-sectional analysis such as this, causal pathways between inflammatory biomarkers and FMD cannot be inferred. The present study identifies an association between an inflammatory marker IL-6 and impaired FMD which serves as a surrogate for endothelial dysfunction. There may be unmeasured confounders that have not been accounted for that could contribute to the association between increased IL-6 levels and impaired FMD. There is always the possibility of measurement errors in FMD, although this should have been minimised by the use of a standardised protocol for FMD measurement. While our study did not show a significant relationship between TNF-α R1 and FMD overall, the sample size of subjects with TNF-α R1 measurements available was small and thus we may not have been able to detect a significant association. However, despite the small sample size, in the subgroup of African Americans, TNF-α R1 was significantly associated with FMD in both the univariate and the complete multivariable models.

Strengths of the present study include the large sample size, use of a standardised protocol for FMD measurement, availability of clinical covariates and use of multivariate analyses in a multi-ethnic community-based sample. Due to the multi-ethnic nature of the cohort, there is increased generalisability of the study. In addition, both men and women were included in this analysis. Further analyses need to be conducted to explore the role of gender with inflammatory biomarkers and FMD.

## Conclusion

Increased level of the pro-inflammatory cytokine IL-6 was associated with impaired endothelial function as assessed by brachial artery FMD in a large multi-ethnic population cohort. The results of the present study suggest that increased circulating IL-6 levels may impair endothelial function, particularly in the brachial artery. Elevated IL-6 concentration may reflect a state that promotes vascular inflammation and development of subclinical atherosclerosis independent of traditional risk factors.

### Key messages

What is already known on this subject?

• Previous studies examining the relationship between inflammatory biomarkers and endothelial dysfunction have shown either weak or no associations of circulating inflammatory biomarkers with flow-mediated dilation.

What this study adds?

• Our study adds to the literature in that it includes data derived from a multi-ethnic community-based cohort that includes non-Caucasians, which to our knowledge has not been previously reported. In addition, we show data divided by race/ethnic group.

How might this impact on clinical practice?

• If further studies confirm these findings, therapies that target IL-6, either indirectly or directly, may be of clinical use.

## Acknowledgments

The authors thank the other investigators, the staff and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.

View Abstract

• ## Supplementary Data

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Files in this Data Supplement:

## Footnotes

• Collaborators Mohammed F Saad, MD.

• Contributors All authors have contributed to the planning, conduct and reporting of the work described in the article. SDW, HNA, ZJ, MRD and SH assume responsibility for the overall content as guarantors.

• Funding This research was supported by contracts N01-HC-95159 through N01-HC-95167 from the National Heart, Lung, and Blood Institute.

• Competing interests None.

• Ethics approval All IRBs at sites involved with the MESA study.

• Provenance and peer review Not commissioned; externally peer reviewed.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.