Abstract
Objective
To examine the relationship of plasma levels of matrix metalloproteinases (MMPs) and tissue inhibitor of metalloproteinase-1 (TIMP-1) with carotid artery characteristics measured by magnetic resonance imaging (MRI) in a cross-sectional investigation among Atherosclerosis Risk in Communities (ARIC) Carotid MRI Study participants.
Methods and Results
A stratified random sample was recruited based on intima-media thickness (IMT) from a previous ultrasound examination. A high-resolution gadolinium-enhanced MRI exam of the carotid artery was performed in 2004–2005 on 1,901 ARIC cohort participants. Multiple carotid wall characteristics including wall thickness, lumen area, calcium area, lipid core and fibrous cap measures were evaluated for associations with plasma MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9 and TIMP-1.
Plasma MMP-1, MMP-3, and MMP-7 were significantly higher among participants in the high IMT group compared to those in the low IMT group. Normalized wall index was independently associated with MMP-3, MMP-7, and TIMP-1. MMP-7 was positively associated with carotid calcification. Mean fibrous cap thickness was significantly higher in individuals with elevated TIMP-1 levels. In addition, TIMP-1 was positively associated with measures of lipid core.
Conclusions
Circulating levels of specific MMPs and TIMP-1 were associated with carotid wall remodeling and structural changes related to plaque burden in the elderly.
Keywords: Atherosclerosis, Carotid, MRI, MMP, TIMP-1, IMT, Luminex, Multianalyte Profiling
Atherosclerosis, the major cause of cardiovascular disease and stroke, is a complex inflammatory process that is characterized by the formation of raised arterial lesions eventually resulting in the narrowing of the lumen.1 Of individuals who suffer an adverse cardiovascular event, as many as 50% have no apparent established traditional risk factor present.2 Recent studies suggest that matrix metalloproteinases (MMPs) play an important role in the early vascular remodeling that accompanies the progression from the initial intima-media thickness (IMT) to the development of a large atheromatous plaque and ultimately to the erosion of the extracellular matrix (ECM) of the fibrous cap which contributes to plaque destabilization and rupture.3 The MMP family consists of more than 24 members of zinc endopeptidases which target a wide variety of substrates, including most of the ECM components that make up the arterial wall.4 MMP-1 and MMP-8 are interstitial collagenases with the ability to cleave the major fibrillar collagens. MMP-2 and MMP-9 are gelatinases that have the ability to digest vascular smooth muscle cell (SMC) basement membranes, collagen, and elastin. MMP-3 is a stromelysin that can degrade a wide variety of substrates including proteoglycans, collagen, and decorin. Belonging to the matrilysins, MMP-7 has a broad range of substrates including collagen and laminin. MMPs are secreted by a variety of cells such as endothelial cells (EC), vascular SMCs, as well as cells involved in the inflammatory cascade.5 The proteolytic activities of MMPs are tightly regulated, and a variety of endogenous inhibitors, including tissue inhibitor of metalloproteinase (TIMPs), are involved.4 TIMP-1, the first discovered and most widely studied of the TIMPs, inhibits most of the active MMPs.
Systemic MMP or TIMP levels may constitute markers of the atherosclerotic process occurring in the vascular tree that are independent of traditional risk factors.6 Adenovirus-mediated over-expression of TIMP-1 in atherosclerosis susceptible apo E-deficient mice significantly reduced atherosclerotic lesions.7 In one of the few direct studies of human arterial tissue, MMP-2, MMP-9, TIMP-1 and TIMP-2 were localized and quantified in carotid endarterectomy (CEA) tissues of normal and atherosclerotic regions.8 The abundance of both MMPs were greater in plaque than in normal segments. TIMPs were less abundant in calcified regions and more abundant in fibrotic and necrotic segments. In a study utilizing both human autopsy and surgical specimens, increased expression of MMP-1 was found in vulnerable atherosclerotic plaques compared to non-lesion areas of the vessel.9 In a small study of 33 patients, peripheral blood levels of MMP-2 and MMP-9 were increased in patients with acute coronary syndrome.10 In a recent case-control study of premature coronary disease in 53 consecutive male patients, plasma levels of MMP-2, MMP-3, MMP-9, TIMP-1, and TIMP-2 were measured, and significant differences were found in all MMPs and TIMPs between patients and controls.11
Although several epidemiological studies have used carotid ultrasound to measure carotid IMT and assess presence or absence of plaque, carotid ultrasound, as performed in ARIC and other studies, cannot reliably quantify plaque and wall volumes, lumen area, or vascular remodeling. MRI can extend the basic ultrasound measurements to assess plaque burden and vascular remodeling. For example, MRI-derived variables such as “normalized wall index” (NWI) can provide a measure of plaque burden which takes into account inherent differences in wall area among vessels of differing diameters. Additionally, MRI provides the opportunity to measure, albeit with less accuracy, other structural components such as the fibrous cap, lipid core, and calcification.
To date, we are not aware of any comprehensive studies such as the current one that has explored the relationships between multiple MMPs and TIMP-1 and specific parameters of carotid artery wall architecture as measured by MRI. Therefore, we have investigated the association of circulating levels of MMPs and TIMP-1 with carotid artery wall characteristics in the large biracial Atherosclerosis Risk in Communities (ARIC) Carotid Magnetic Resonance Imagery (MRI) Study cohort.
METHODS
Study Design and Study Participants
The ARIC Study is a prospective investigation of atherosclerosis involving 15,792 African American and Caucasian men and women aged 45 to 64 years at recruitment (1987–1989). Participants underwent a baseline and up to 3 follow-up visits through 1998. A detailed description of the ARIC study design and methods has been published elsewhere.12 The ARIC Carotid MRI Study was conducted between 2004 and 2005 as a cross-sectional sub-study. The participants were selected from the larger ARIC Study cohort based on results of the last ultrasound examinations (Visits 3 and 4, 1993–1998). In order to increase the prevalence of informative plaques while maintaining the ability to make population-based inferences, a stratified sampling plan was used. The goal was to recruit 1,200 participants with high values of maximum carotid artery wall thickness (maximum over 6 sites: common, bifurcation, internal carotid arteries of the left and right sides) and 800 participants randomly sampled from the remainder of the cohort below this cut point. This carotid artery IMT cut point was determined for each field center based on the IMT distribution specific to that site, and field-center-specific cut points of carotid IMT were adjusted over the recruitment period to approximately achieve this goal.
Recruitment lists were provided to the field centers, which sampled from above and below the IMT cut point as indicated by the sampling plan. Ineligibility criteria for the Carotid MRI Study included standard contraindications to the MRI exam or to the contrast agent, carotid revascularization on either side for the low IMT group or on the imaging-selected side for the high IMT group. Of the total of 4,307 persons who were contacted and invited to participate in the study, 1,404 refused, 837 were ineligible, and 2,066 participated. The overall recruitment rates were 40%, 48%, 51%, and 53% of all persons contacted from the 4 centers, with little difference in rates between high and low IMT groups. In addition, selected participants also received in-person interviews and a physical examination.
Of these 2,066 ARIC cohort members who participated in the Carotid MRI Study, 1,901 had a complete MRI exam, and 1,769, of the 1,901, had a sufficient quality of MRI scans and adherence to the MRI protocol to be included. Additionally, 91 participants with 1 or more missing values for any of the MMP or TIMP-1 measurements were excluded leaving 1,678 subjects in the final analyses. The average age of these 1,678 participants was 70 years and included 1,304 Caucasians (632 women and 672 men) and 374 African-Americans (220 women and 154 men).
Participant Examination
Protocols for lipoprotein, fasting glucose, blood pressure (BP), and height and weight measurements were identical at the ARIC cohort baseline examination and the Carotid MRI Sub-study examination conducted 18 years later. Cigarette smoking was ascertained from an interview and categorized as “current,” “former,” or “never.” Prescription drug use was obtained by self-report. Fasting blood samples were collected and assayed for total cholesterol and high-density lipoprotein (HDL) cholesterol. Low-density lipoprotein (LDL) cholesterol was calculated according to the Friedewald formula. We assessed C-reactive protein (CRP) levels by the immunoturbidimetric CRP-Latex (II) high-sensitivity assay from Denka Seiken (Tokyo, Japan) performed according to the manufacturer's protocol using an Olympus AU400e automated chemistry analyzer (Olympus America Inc., Melville, NY). Resting BP was determined 3 times using a random-zero sphygmomanometer, and the mean of the last 2 measurements was used. Hypertension was defined as a systolic BP ≥140 mm Hg, diastolic BP ≥90 mm Hg, or use of antihypertensive medication during the previous 2 weeks. Diabetes mellitus (DM) was defined as a fasting glucose level of at least 126 mg/dL, a non-fasting glucose level of at least 200 mg/dL, or a self-reported history of physician-diagnosed DM or treatment for DM. Previous history of cardiovascular disease (CVD) included adjudicated myocardial infarction (MI), coronary heart disease (CHD) death, stroke, and/or revascularization procedure.13 The study was approved by the institutional review committees of all participating centers and all participants provided informed consent.
MRI Protocol
MRI studies were performed on a 1.5T scanner (GE Medical Systems, Milwaukee, WI at 3 field centers; Siemens Medical Solutions, Ehrlangen, Germany at one field center) equipped with a bilateral 4-element phased array carotid coil (Machnet, The Netherlands). A 3-dimensional time-of-flight magnetic resonance angiogram (MRA) was acquired through both carotid bifurcations. Detailed black blood MRI (BBMRI) images were then acquired through the extracranial carotid bifurcation known to have a thicker maximum wall by the most recent ultrasound study, unless the contralateral carotid bifurcation wall appeared thicker on the MRA to the technologist. BBMRI imaging was achieved using a cardiac-gated, 2-dimensional double inversion recovery fast spin echo sequence with the inversion time set to suppress the signal of blood. Each participant received an intravenous injection of gadodiamide (Omniscan, GE Amersham), 0.1 mmol/kg body weight, using a power injector. Sixteen transverse T1-weighted BBMRI images (acquired resolution, 0.51×0.58×2mm3; total longitudinal coverage, 3.2cm) were oriented perpendicular to the vessel and centered at the thickest part of the internal or common carotid artery wall. These 16 slices were acquired 5 minutes after the injection through the thicker carotid artery using a 2-dimensional double inversion recovery fast spin echo sequence with the inversion time set to suppress the signal of the contrast-enhanced blood pool.
Image Analysis
MRI images were analyzed by 7 specially trained readers, blinded to the clinical and laboratory characteristics of the study population. All exams were assessed for image quality and protocol adherence, and exams that failed were not analyzed.
Slices were numbered 1 through 16 from proximal to distal, and only 8 slices, centered around the slice with the thickest wall, were analyzed. The plaque components were analyzed on the postcontrast BBMRI series based on the ability of gadolinium enhancement to delineate and enable quantitative size measurements of the fibrous cap,14 and contours were drawn to delineate the outer wall, lumen, lipid core, and calcification. The total vessel area included lumen, intima, media and adventitia. Wall area was calculated as the difference between total vessel area and lumen area. The derived MRI variable Normalized Wall Index (NWI) was calculated by dividing the wall area by the total vessel area. Reliability coefficients (R) for the aforementioned MRI variables were obtained from an internal reliability study and have been published previously.15 Generally, reliability for most MRI variables varied from excellent (R>0.75) to fair-to-good (0.4<R<0.75). However, the limited resolution of the 1.5T MRI may have contributed to the lower reliability of measurements related to relatively small structural features such as the fibrous cap thickness. Although, reliability based on repeated readings was fair to good for cap thickness measures (R=0.60), reliability based on repeated scans was poor for cap thickness (R=0.38). A detailed description of the image analysis methods for the ARIC carotid MRI study has been published previously.15
Multianalyte profiling (MAP) assay
Fluorokine MAP allows the simultaneous measurement of multiple biomarkers from small sample volumes. In the current study, 6 MMP analytes were measured utilizing an assay from R&D Systems, Inc. (Minneapolis, MN) according to manufacturer's protocol with the following minor modifications. Due to differences in the plasma levels of the MMPs and the sensitivity of the assay, two different dilutions were utilized. Panel A measured MMP-1, MMP-7, and MMP-8 which required a 2-fold dilution, while Panel B measured MMP-2, MMP-3, and MMP-9 which required a 10-fold dilution. MMP control materials were obtained from R&D, and appropriate mixtures for Panel A and Panel B were run for each assay. Duplicates were run for all test samples, controls, and standards. Bio-Plex Manager 4.1 software (BioRad Labs) was used for data acquisition. TIMP-1 was measured with a solid-phase sandwich ELISA technique (R&D Systems) according to the manufacturer's protocol. Reliability coefficients for MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9, and TIMP-1 were 0.87, 0.61, 0.66, 0.82, 0.79, 0.78, and 0.60, respectively based on blinded replicate pairs of 60 participants. Intra- and inter-assay coefficient of variation for MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9, and TIMP-1 were 7.0 and 8.6%, 3.4 and 5.9%, 3.9 and 6.2%, 5.1 and 12.6%, 7.6 and 11.5%, 3.6 and 5.6%, and 4.7 and 6.9%, respectively. MMPs were measured as the total of all forms including pro-, mature-, and TIMP-complexed-forms; total TIMP-1 was also measured as both the free- and MMP-complexed-forms.
Statistical Analysis
All analyses were based on methods appropriate for a stratified random sampling. In particular, all analyses were weighted by the inverse of the sampling fractions in the 8 sampling strata (4 field centers × 2 IMT groups). The sampling fractions were based on those persons screened for participation. Those who actually participated were analyzed as a domain for calculating variances and confidence intervals of estimators. Analyses were conducted using SAS version 9.1 for descriptive statistics or SUDAAN for domain analysis. Finite population correction factors were not applied. Tests of differences in weighted means or proportions between groups were from weighted linear or logistic regression models that accounted for the sampling. Adjusted means and proportions by sub-group of interest were calculated using SUDAAN REGRESSION for continuous variables and LOGISTIC for dichotomous variables, with predicted values calculated as sample means of the adjusting variables. Wall thickness and wall volume were analyzed in the full data set of 1,678 participants. Due to the resolution constraints of the MRI scan, we restricted consideration of lipid core to the 1,156 participants whose maximum wall thickness was ≥1.5 mm (weighted percentage=62%). Only 4 lipid cores were excluded using this cut point. Measures of lipid core volume and area, and fibrous cap thickness were analyzed as continuous variables among those 554 participants with a lipid core. An additional analysis considered correlates of lipid core presence as a dichotomous variable. Standardized regression coefficients are presented for linear and logistic regression models, standardizing by 1-standard deviation of exposure and outcome (for continuous outcomes) with adjustment for age, race, and gender.
RESULTS
We investigated the relationship between circulating levels of MMPs and TIMP-1 and a number of traditional cardiovascular risk factors, as shown in Table 1. With the exception of MMP-9, all measured MMPs and TIMP-1 were positively associated with age. Interestingly, MMP-1 and MMP-2 were negatively associated with LDL-cholesterol. We also found that circulating levels of MMP-1, MMP-8, MMP-9, and TIMP-1 were negatively associated with HDL-cholesterol. Furthermore, MMP-8 and TIMP-1 were positively related to body mass index (BMI) as well as the inflammatory marker CRP. Elevated plasma levels of MMP-7 and TIMP-1 were associated with a higher incidence of hypertension, and MMP-1 and MMP-7 were similarly increased with incident diabetes. Individuals who used statins were more likely to have elevated plasma levels of MMP-1, MMP-2 and MMP-7. No obvious relationships were observed between circulating levels of MMP-s/TIMP-1 and smoking status or family history of CVD. However, individuals with a prior self-history of CVD were more likely to have elevated (3rd tertile) MMP-3 plasma levels.
Table 1.
Weighted means (SE) or proportion (SE) of Risk factors by weighted tertiles of each MMP/TIMP variable, and p-value for test for any differences between tertiles, adjusted by age, race, gender
| Risk Factors* | MMP/TIMP Variables | 1st Tertile | 2nd Tertile | 3rd Tertile | P-Value† |
|---|---|---|---|---|---|
| Age (years) (n=1678) | MMP-1 | 69.72 (0.282) | 70.32 (0.276) | 70.96 (0.307) | 0.0149 |
| MMP-2 | 69.10 (0.283) | 70.24 (0.277) | 71.62 (0.287) | <0.0001 | |
| MMP-3 | 69.13 (0.291) | 70.32 (0.282) | 71.51 (0.309) | <0.0001 | |
| MMP-7 | 69.35 (0.287) | 69.81 (0.278) | 71.81 (0.279) | <0.0001 | |
| MMP-8 | 69.65 (0.281) | 70.24 (0.298) | 71.08 (0.276) | 0.0013 | |
| MMP-9 | 70.61 (0.310) | 70.07 (0.276) | 70.26 (0.299) | 0.4188 | |
| TIMP-1 | 68.75 (0.238) | 70.21 (0.291) | 72.00 (0.300) | <0.0001 | |
| LDL Cholesterol (mg/dl) (n=1650) | MMP-1 | 117.65 (1.999) | 114.99 (2.092) | 108.55 (1.824) | 0.0030 |
| MMP-2 | 117.57 (1.968) | 114.09 (1.990) | 109.94 (1.921) | 0.0249 | |
| MMP-3 | 109.80 (2.039) | 116.31 (1.931) | 115.53 (2.078) | 0.0587 | |
| MMP-7 | 114.87 (1.930) | 115.30 (2.031) | 111.43 (1.892) | 0.3002 | |
| MMP-8 | 115.38 (1.909) | 115.16 (2.128) | 111.00 (1.768) | 0.1605 | |
| MMP-9 | 114.84 (2.024) | 116.65 (1.736) | 110.10 (2.242) | 0.0809 | |
| TIMP-1 | 114.59 (1.983) | 113.46 (1.899) | 113.52 (2.084) | 0.9051 | |
| HDL Cholesterol (mg/dl) (n=1678) | MMP-1 | 50.93 (0.762) | 51.45 (0.830) | 48.64 (0.710) | 0.0176 |
| MMP-2 | 49.81 (0.745) | 49.83 (0.748) | 51.43 (0.820) | 0.2395 | |
| MMP-3 | 49.29 (0.875) | 51.01 (0.717) | 50.78 (0.825) | 0.3128 | |
| MMP-7 | 50.45 (0.738) | 51.02 (0.770) | 49.60 (0.811) | 0.4344 | |
| MMP-8 | 51.50 (0.806) | 51.15 (0.747) | 48.38 (0.725) | 0.0039 | |
| MMP-9 | 52.41 (0.817) | 50.04 (0.752) | 48.59 (0.791) | 0.0034 | |
| TIMP-1 | 52.29 (0.769) | 51.14 (0.839) | 47.63 (0.669) | <0.0001 | |
| BMI (kg/m2) (n=1677) | MMP-1 | 28.34 (0.254) | 28.24 (0.273) | 28.56 (0.265) | 0.6869 |
| MMP-2 | 28.91 (0.281) | 28.18 (0.246) | 28.04 (0.244) | 0.0515 | |
| MMP-3 | 28.66 (0.300) | 28.43 (0.243) | 28.04 (0.272) | 0.3131 | |
| MMP-7 | 28.17 (0.252) | 28.52 (0.266) | 28.44 (0.260) | 0.6066 | |
| MMP-8 | 27.28 (0.240) | 28.52 (0.254) | 29.37 (0.267) | <0.0001 | |
| MMP-9 | 28.30 (0.283) | 28.53 (0.250) | 28.31 (0.260) | 0.7866 | |
| TIMP-1 | 27.50 (0.247) | 28.34 (0.266) | 29.31 (0.262) | <0.0001 | |
| Log(CRP) (n=1669) | MMP-1 | 0.78 (0.050) | 0.70 (0.052) | 0.69 (0.058) | 0.4397 |
| MMP-2 | 0.88 (0.055) | 0.74 (0.050) | 0.55 (0.052) | 0.0001 | |
| MMP-3 | 0.62 (0.056) | 0.77 (0.053) | 0.78 (0.056) | 0.1138 | |
| MMP-7 | 0.77 (0.054) | 0.74 (0.053) | 0.65 (0.050) | 0.2420 | |
| MMP-8 | 0.49 (0.049) | 0.69 (0.052) | 1.00 (0.053) | <0.0001 | |
| MMP-9 | 0.73 (0.053) | 0.63 (0.055) | 0.80 (0.052) | 0.0841 | |
| TIMP-1 | 0.53 (0.054) | 0.69 (0.049) | 0.94 (0.053) | <0.0001 | |
| Hypertensive (n=1663) | MMP-1 | 60.0% (2.60%) | 62.2% (2.59%) | 66.1% (2.64%) | 0.2657 |
| MMP-2 | 63.1% (2.64%) | 59.2% (2.64%) | 65.9% (2.55%) | 0.1877 | |
| MMP-3 | 60.3% (2.82%) | 60.5% (2.64%) | 67.3% (2.59%) | 0.1294 | |
| MMP-7 | 57.7% (2.65%) | 60.5% (2.62%) | 70.6% (2.54%) | 0.0016 | |
| MMP-8 | 59.8% (2.60%) | 61.4% (2.61%) | 67.3% (2.58%) | 0.0994 | |
| MMP-9 | 61.9% (2.65%) | 61.5% (2.60%) | 65.0% (2.78%) | 0.6415 | |
| TIMP-1 | 57.6% (2.62%) | 63.0% (2.63%) | 67.8% (2.59%) | 0.0239 | |
| Diabetes (n=1661) | MMP-1 | 16.3% (1.77%) | 22.4% (2.14%) | 29.3% (2.54%) | 0.0003 |
| MMP-2 | 21.3% (2.18%) | 22.4% (2.17%) | 23.1% (2.13%) | 0.8403 | |
| MMP-3 | 22.4% (2.43%) | 21.7% (2.13%) | 22.6% (2.27%) | 0.9492 | |
| MMP-7 | 17.0% (2.01%) | 24.0% (2.27%) | 25.8% (2.21%) | 0.0133 | |
| MMP-8 | 22.7% (2.23%) | 19.3% (2.06%) | 24.5% (2.21%) | 0.2220 | |
| MMP-9 | 20.9% (2.24%) | 20.3% (2.14%) | 25.2% (2.35%) | 0.2731 | |
| TIMP-1 | 20.0% (2.15%) | 21.8% (2.16%) | 24.8% (2.21%) | 0.2940 | |
| Statin use (n=1656) | MMP-1 | 24.2% (2.12%) | 34.7% (2.50%) | 42.9% (2.77%) | <0.0001 |
| MMP-2 | 27.6% (2.39%) | 39.4% (2.62%) | 34.1% (2.46%) | 0.0051 | |
| MMP-3 | 37.7% (2.84%) | 32.5% (2.48%) | 31.2% (2.41%) | 0.2441 | |
| MMP-7 | 29.5% (2.45%) | 33.5% (2.51%) | 38.4% (2.56%) | 0.0499 | |
| MMP-8 | 30.1% (2.41%) | 33.8% (2.52%) | 37.5% (2.57%) | 0.1196 | |
| MMP-9 | 29.3% (2.37%) | 34.6% (2.50%) | 37.7% (2.79%) | 0.0765 | |
| TIMP-1 | 34.8% (2.54%) | 32.4% (2.49%) | 34.0% (2.49%) | 0.7820 | |
| Self-History of CVD (Stroke or CHD) (n=1678) | MMP-1 | 9.7% (1.50%) | 10.5% (1.46%) | 12.2% (1.66%) | 0.5373 |
| MMP-2 | 9.3% (1.58%) | 9.1% (1.37%) | 13.7% (1.67%) | 0.0528 | |
| MMP-3 | 10.5% (1.96%) | 7.3% (1.23%) | 13.8% (1.56%) | 0.0037 | |
| MMP-7 | 9.4% (1.50%) | 10.1% (1.52%) | 13.0% (1.65%) | 0.2333 | |
| MMP-8 | 7.9% (1.36%) | 12.6% (1.70%) | 11.8% (1.49%) | 0.0834 | |
| MMP-9 | 9.2% (1.25%) | 11.0% (1.60%) | 12.7% (1.97%) | 0.2958 | |
| TIMP-1 | 9.7% (1.58%) | 9.2% (1.44%) | 13.3% (1.62%) | 0.1138 | |
| Family History of CVD (Stroke or CHD) (n=1663) | MMP-1 | 56.3% (2.55%) | 55.2% (2.63%) | 58.5% (2.80%) | 0.6784 |
| MMP-2 | 54.1% (2.73%) | 56.2% (2.60%) | 59.7% (2.61%) | 0.3373 | |
| MMP-3 | 55.2% (2.84%) | 55.3% (2.66%) | 59.3% (2.75%) | 0.5090 | |
| MMP-7 | 52.7% (2.67%) | 57.7% (2.64%) | 59.6% (2.61%) | 0.1630 | |
| MMP-8 | 53.5% (2.68%) | 59.2% (2.61%) | 57.3% (2.61%) | 0.3136 | |
| MMP-9 | 60.6% (2.68%) | 54.4% (2.67%) | 54.8% (2.82%) | 0.2011 | |
| TIMP-1 | 60.5% (2.58%) | 53.8% (2.66%) | 55.5% (2.66%) | 0.1784 | |
| Current Smoker (n=1656) | MMP-1 | 6.5% (1.22%) | 8.9% (1.53%) | 7.7% (1.43%) | 0.2851 |
| MMP-2 | 6.6% (1.26%) | 9.8% (1.60%) | 6.5% (1.27%) | 0.4630 | |
| MMP-3 | 7.8% (1.46%) | 6.9% (1.25%) | 8.2% (1.57%) | 0.5827 | |
| MMP-7 | 6.6% (1.31%) | 7.4% (1.33%) | 9.1% (1.50%) | 0.5753 | |
| MMP-8 | 7.1% (1.34%) | 5.9% (1.23%) | 10.0% (1.58%) | 0.2233 | |
| MMP-9 | 4.9% (1.18%) | 7.6% (1.43%) | 10.3% (1.72%) | 0.1661 | |
| TIMP-1 | 5.5% (1.14%) | 8.5% (1.46%) | 9.2% (1.58%) | 0.3397 |
n=number of included participants
P-value for test of differences in means
Supplemental Table 1 shows the age-adjusted, race- and gender-specific plasma levels of MMPs and TIMP-1. Plasma levels of MMP-2 and MMP-3 were significantly higher in men than women of either race. In contrast, plasma levels of MMP-7 were lower overall in men than women. Furthermore, African Americans had increased plasma levels of MMP-7, MMP-8, and MMP-9 when compared to those in Caucasians. Strikingly, circulating levels of MMP-9 were elevated nearly 2-fold in African Americans. In addition to circulating levels of MMPs, a number of traditional cardiovascular risk factors also differed significantly by ethnicity and gender. C-reactive protein (CRP) was increased in women compared to men and higher overall in African Americans than Caucasians. The prevalence of both hypertension and diabetes was also higher overall in African Americans than Caucasians.
The simultaneous measurement of a number of different MMPs in this large cohort study enabled us to investigate potentially significant inter-relationships, which are summarized in Table 2. Generally, correlations between the various MMPs and TIMP-1 were low to moderate (Pearson R varied from −0.008 [between TIMP-1 and MMP-9] to 0.352 [between TIMP-1 and MMP-7]). TIMP-1 showed significant positive correlations with all MMPs measured with the exception of MMP-9. In contrast, MMP-9 did not correlate with any of the other MMPs measured except for a significant positive correlation with MMP-8.
TABLE 2.
Weighted Pearson correlation coefficient between circulating levels of different MMPs and TIMP-1
| MMP-1 | MMP-2 | MMP-3 | MMP-7 | MMP-8 | MMP-9 | TIMP-1 | |
|---|---|---|---|---|---|---|---|
| MMP-1 | 0.044 | 0.150† | 0.138‡ | 0.162* | −0.017 | 0.102‡ | |
| MMP-2 | 0.140‡ | 0.137‡ | 0.031* | 0.017 | 0.262† | ||
| MMP-3 | 0.154‡ | 0.062‡ | −0.025 | 0.133‡ | |||
| MMP-7 | 0.022 | 0.037 | 0.352‡ | ||||
| MMP-8 | 0.214* | 0.095‡ | |||||
| MMP-9 | −0.008 | ||||||
| TIMP-1 |
P<0.05
P<0.001
P<0.0001
The weighted means or proportions of traditional cardiovascular risk factors and MMPs and TIMP-1, adjusted by age, race, and gender are shown in Table 3. The results are stratified into high and low IMT groups, based upon ultrasound exams for participant selection as described in the “Methods” section. Systolic BP, BMI, smoking (both current and former), hypertension, and lipid-lowering medication (statin) use were all greater in the high IMT group. Furthermore, circulating levels of MMP-1, MMP-3, and MMP-7 were significantly elevated in the high IMT participants.
Table 3.
Weighted mean (SE), or proportions of traditional risk factors and MMPs and TIMP-1 levels (mean [SE]) among MRI participants stratified by imputed IMT, adjusted by age, gender and race.
| Characteristics | Low IMT group | High IMTgroup | P† |
|---|---|---|---|
| Number of included participants (N) | 553 | 1125 | |
| LDL Cholesterol (mg/dl) | 113.97 (1.264) | 113.24 (1.949) | 0.7511 |
| HDL Cholesterol (mg/dl) | 50.43 (0.511) | 49.94 (0.664) | 0.5401 |
| Systolic Blood Pressure (mmHg) | 125.00 (0.643) | 128.63 (1.031) | 0.0028 |
| Diastolic Blood Pressure (mmHg) | 67.01 (0.349) | 65.86 (0.519) | 0.0645 |
| BMI (kg/m2) | 28.29 (0.166) | 28.92 (0.219) | 0.0154 |
| Log (CRP) | 0.72 (0.035) | 0.75 (0.046) | 0.5641 |
| Current Smoker | 7.0% | 12.6% | 0.0004‡ |
| Former Smoker | 41.9% | 47.3% | |
| Hypertensive | 60.9% | 73.7% | <0.0001 |
| Hypertensive medication use | 53.8% | 68.3% | <0.0001 |
| Diabetes | 21.6% | 25.8% | 0.1101 |
| Diabetes Medication use | 13.4% | 17.1% | 0.0959 |
| Statin use | 31.5% | 46.6% | <0.0001 |
| Anti-inflammatory Medication use | 77.5% | 83.3% | 0.0146 |
| MMP-1 (ng/ml) | 0.172 (0.008) | 0.210 (0.011) | 0.0040 |
| MMP-2 (ng/ml) | 299.51 (4.30) | 313.18 (6.35) | 0.0849 |
| MMP-3 (ng/ml) | 6.30 (0.14) | 7.30 (0.34) | 0.0070 |
| MMP-7 (ng/ml) | 2.47 (0.057) | 2.92 (0.10) | 0.0001 |
| MMP-8 (ng/ml) | 0.452 (0.023) | 0.457 (0.014) | 0.8178 |
| MMP-9 (ng/ml) | 21.02 (0.39) | 20.96 (0.44) | 0.9261 |
| TIMP-1 (ng/ml) | 111.53 (1.07) | 114.64 (1.57) | 0.1103 |
*Data presented as mean (standard error of the mean) or percentages.
P-value for test of differences in means.
P-value for test if 2 groups are different in current smoking OR former smoking. Thus, it is a 2-degrees-of-freedom test.
We examined the relationship of individual MMPs and TIMP-1 with various carotid wall measures as determined by MRI (Table 4). Circulating levels of MMP-7 (p=0.0394) and TIMP-1 (p=0.0192) were positively and statistically significantly associated with total wall volume after adjustment for major cardiovascular risk factors. In order to examine possible associations of circulating levels of MMPs and TIMP-1 with plaque burden, we included the derived MRI variable “normalized wall index,” or NWI, in our analysis as previously described by Saam et al.16 NWI is calculated as the wall area divided by the total vessel area (wall area + lumen area). NWI takes into account inherent differences in wall area among vessels of differing diameters and therefore may provide a more accurate measure of plaque burden than measurement of stenosis alone.17 Plasma levels of MMP-3, MMP-7, and TIMP-1 were positively associated with NWI after adjustment for major cardiovascular risk factors. To further investigate the relationship of MMPs and TIMP-1 with various carotid MRI variables, we focused our analyses on those participants who had a maximum carotid wall thickness of greater than 1.5 mm and presence of atherosclerotic plaque as determined by the presence of lipid core (Table 5). Regression analysis showed that circulating levels of TIMP-1 were independently and positively associated with fibrous cap thickness in a fully adjusted model (p=0.0439). Additional analysis in which fibrous cap thickness was treated as a dichotomized variable (i.e. above or below the weighted median) showed that individuals with plasma TIMP-1 levels in the 3rd tertile had an OR of 1.42 (95%CI; 0.79 – 2.55) for increased fibrous cap thickness compared to those individuals with TIMP-1 levels in the 1st tertile.
Table 4.
Weighted means (SE) or proportion (SE) of MRI variables by weighted tertiles of each MMP/TIMP variable, and p-value for test for any differences between tertiles, adjusted by age, race, gender, smoking, hypertension, CRP, diabetes, total cholesterol, and HDL cholesterol, on all participants
| MRI Variables | CLDA Variables | 1st Tertile | 2nd Tertile | 3rd Tertile | P-Value |
|---|---|---|---|---|---|
| Total Wall Volume (n=1678)* | MMP-1 | 4.065 (0.073) | 4.131 (0.080) | 4.210 (0.087) | 0.4414 |
| MMP-2 | 4.118 (0.084) | 4.088 (0.078) | 4.195 (0.083) | 0.6232 | |
| MMP-3 | 4.120 (0.080) | 4.038 (0.079) | 4.245 (0.090) | 0.2174 | |
| MMP-7 | 3.984 (0.069) | 4.193 (0.094) | 4.231 (0.078) | 0.0394 | |
| MMP-8 | 4.126 (0.078) | 4.136 (0.077) | 4.140 (0.086) | 0.9921 | |
| MMP-9 | 4.150 (0.081) | 4.216 (0.086) | 4.034 (0.074) | 0.2500 | |
| TIMP-1 | 3.963 (0.075) | 4.278 (0.092) | 4.161 (0.076) | 0.0192 | |
| Normalized Wall Index (NWI) (n=1678)* | MMP-1 | 0.432 (0.007) | 0.432 (0.007) | 0.450 (0.007) | 0.1406 |
| MMP-2 | 0.438 (0.007) | 0.435 (0.007) | 0.441 (0.007) | 0.8164 | |
| MMP-3 | 0.431 (0.007) | 0.426 (0.007) | 0.457 (0.008) | 0.0080 | |
| MMP-7 | 0.426 (0.007) | 0.438 (0.007) | 0.450 (0.007) | 0.0587 | |
| MMP-8 | 0.433 (0.007) | 0.438 (0.007) | 0.444 (0.008) | 0.5688 | |
| MMP-9 | 0.433 (0.007) | 0.443 (0.007) | 0.437 (0.008) | 0.5807 | |
| TIMP-1 | 0.419 (0.006) | 0.448 (0.007) | 0.447 (0.007) | 0.0040 |
n=number of included participants
Table 5.
Weighted means (SE) or proportion (SE) of MRI variables by weighted tertiles of each MMP/TIMP variable, and p-value for test for any differences between tertiles, adjusted by age, race, gender, smoking, hypertension, CRP, diabetes, total cholesterol, and HDL cholesterol, and restricted to participants with Max Wall Thickness >1.5 and Lipid Core=1.
| MRI Variables | CLDA Variables | 1st Tertile | 2nd Tertile | 3rd Tertile | P-Value* |
|---|---|---|---|---|---|
| Mean Fibrous Cap Thickness (mm) (n=544)* | MMP-1 | 0.682 (0.028) | 0.642 (0.031) | 0.700 (0.028) | 0.3684 |
| MMP-2 | 0.678 (0.029) | 0.673 (0.033) | 0.678 (0.026) | 0.9918 | |
| MMP-3 | 0.682 (0.032) | 0.683 (0.033) | 0.668 (0.026) | 0.9254 | |
| MMP-7 | 0.684 (0.034) | 0.640 (0.026) | 0.701 (0.026) | 0.2287 | |
| MMP-8 | 0.673 (0.026) | 0.652 (0.037) | 0.700 (0.024) | 0.5126 | |
| MMP-9 | 0.663 (0.034) | 0.664 (0.028) | 0.702 (0.028) | 0.6090 | |
| TIMP-1 | 0.626 (0.024) | 0.669 (0.032) | 0.718 (0.027) | 0.0439 | |
| Maximum Calcium Area (mm2) (n=545)* | MMP-1 | 0.032 (0.003) | 0.038 (0.004) | 0.044 (0.006) | 0.0642 |
| MMP-2 | 0.042 (0.009) | 0.037 (0.003) | 0.037 (0.004) | 0.8757 | |
| MMP-3 | 0.034 (0.003) | 0.036 (0.004) | 0.044 (0.006) | 0.3264 | |
| MMP-7 | 0.027 (0.002) | 0.045 (0.008) | 0.043 (0.004) | 0.0002 | |
| MMP-8 | 0.045 (0.008) | 0.033 (0.003) | 0.038 (0.004) | 0.2209 | |
| MMP-9 | 0.042 (0.006) | 0.039 (0.004) | 0.033 (0.003) | 0.3015 | |
| TIMP-1 | 0.035 (0.004) | 0.042 (0.007) | 0.038 (0.003) | 0.4951 | |
| Maximum Lipid Core Area (mm2) (n=554)* | MMP-1 | 0.102 (0.011) | 0.115 (0.010) | 0.109 (0.010) | 0.6936 |
| MMP-2 | 0.098 (0.007) | 0.110 (0.009) | 0.115 (0.013) | 0.3876 | |
| MMP-3 | 0.110 (0.009) | 0.097 (0.008) | 0.115 (0.011) | 0.3724 | |
| MMP-7 | 0.099 (0.008) | 0.121 (0.013) | 0.105 (0.011) | 0.3207 | |
| MMP-8 | 0.112 (0.012) | 0.101 (0.009) | 0.111 (0.011) | 0.7276 | |
| MMP-9 | 0.101 (0.010) | 0.1348 (0.013) | 0.0922 (0.007) | 0.0062 | |
| TIMP-1 | 0.089 (0.008) | 0.1325 (0.014) | 0.0990 (0.008) | 0.0146 |
n=number of included participants
Plasma levels of MMP-1 and MMP-7 were positively associated with carotid artery calcification, although the association for MMP-1 did not quite reach significance at the p<0.05 level. Interestingly, MMP-9 showed a moderate inverse association with carotid artery calcification. Both MMP-9 (p=0.0062) and TIMP-1 (p=0.0146) levels were associated with measures of lipid core, including maximum lipid core area and maximum lipid core volume. However, MMP-9 and TIMP-1 were significantly elevated in the 2nd tertile only. Additional analyses showed that circulating levels of TIMP-1 were associated with increased odds for presence of lipid core in a model adjusted for age, gender, and race, although odds were somewhat attenuated in a fully adjusted model (Table 6).
Table 6.
Adjusted odds ratios for lipid core presence according TIMP-1 levels modeled in tertiles
| Model 1 | Model 2 | |||||
|---|---|---|---|---|---|---|
| Variable | Odds Ratio | 95% CI | P | Odds Ratio | 95% CI | P |
| TIMP-1 | ||||||
| 2nd tertile vs. 1st tertile | 1.49 | 1.02 – 2.17 | 0.039 | 1.37 | 0.93 – 2.01 | 0.113 |
| 3rd tertile vs. 1st tertile | 1.54 | 1.05 – 2.24 | 0.026 | 1.41 | 0.95 – 2.10 | 0.092 |
Model 1: Adjusted for age, gender and race
Model 2: Adjusted for variables included in Model 1 and smoking, hypertension, diabetes, total cholesterol, HDL cholesterol, and CRP.
DISCUSSION
Plasma levels of MMPs and TIMP-1 were associated with a number of conventional cardiovascular risk factors (Table 1). In particular, we found positive associations with age for most MMPs and TIMP-1, with the exception of MMP-9. In one of the few studies designed to investigate the effect of age on MMPs and TIMPs, a small cross-sectional study of 27 asymptomatic individuals aged 20–90 examined MMP-2, MMP-7, MMP-8, MMP-9, and TIMP-1, TIMP-2, and TIMP-4 plasma concentrations.18 The authors reported increased levels of MMP-2, MMP-7, and all TIMPs, and reported no change in MMP-8 levels with age, the latter result being the only deviation from the current finding. Plasma levels of MMP-8 and TIMP-1 were positively associated with BMI as well as the inflammatory marker, CRP. Furthermore, hypertensive subjects and diabetics were more likely to have elevated levels of MMP-1, MMP-7 and TIMP-1. Overall, our findings show that circulating levels of MMPs and TIMP-1 were associated with an adverse cardiovascular risk profile, and therefore, are supportive of the current belief that MMPs play a critical role in the etiology of atherosclerosis and CVD.
The ARIC Carotid MRI study also provided a unique opportunity to examine the effect of race and gender on plasma levels of MMPs and TIMP-1 in a large biracial population. We found that plasma levels of MMP-2 and MMP-3 were significantly increased, and MMP-7 was decreased in men compared to women (Supplemental Table 1). A similar finding was observed in a case-control study of 387 myocardial infarction patients in which the MMP-3 serum concentrations for both case and control subjects were almost double for men compared to women.19 Furthermore, in our biracial study cohort, MMP-7, MMP-8, and MMP-9 levels were elevated in African Americans compared to Caucasians.
A number of studies have shown that possible pleiotropic effects of statins may involve the down-regulation of MMP expression. Cerivastatin inhibited inducible MMP-1, MMP-3, and MMP-9 secretion from human SMCs.20 In a study of 32 patients with CAD, 14 weeks of simvastatin treatment significantly reduced plasma MMP-9 concentration.21 It has been hypothesized that statins may elicit some of their atheroprotective functions by decreasing MMP levels thus leading to stabilization of the plaques with or without lesion regression. Our results, which show that MMP-1 and MMP-2 were negatively associated with LDL-cholesterol and individuals with elevated plasma levels of MMP-1 and MMP-2 were more likely to be on statin therapy (Table 1), may appear to contradict the findings from these previous studies. However, our findings may be confounded because high-risk patients (i.e., those with coronary heart disease) were more likely to be taking a statin. In addition, it is important to note that the ARIC study was not designed to investigate possible effects of statin therapy on circulating MMP levels.
We investigated the inter-relationships among MMPs and TIMP-1 plasma concentrations (Table 2). Significant correlations may result from coordinated gene expression or could represent a commonality of the tissues and cell types responsible for production of the MMPs that are mutually up- or down-regulated in response to shared agonists or antagonists.22 We found MMP-8 and MMP-9 levels were correlated with one another (r=0.214, p <0.0001). A significant relationship between MMP-8 and MMP-9 was also found in a study of CEA specimens,23 and strong correlations between those MMPs were noted in a study of adolescents.24 MMP-8 and MMP-9 are both synthesized by differentiating granulocytes, stored in circulating neutrophils, and released following neutrophil activation,25 a mutuality that could partially explain their association. TIMP-1 was significantly positively correlated with all MMPs except MMP-9; among these significant correlations, the association with MMP-8 was the weakest. The strongest correlation (r=0.352) was found between TIMP-1 and MMP-7 suggesting that their regulation may be closely linked. In the present study, both MMP-7 and TIMP-1 concentrations were significantly associated with total wall volume, maximum wall thickness at the lipid core, and NWI suggesting shared functional associations possibly related to their correlation.
Circulating levels of MMP-1, MMP-3, and MMP-7 were elevated among participants in the high IMT group as determined by ultrasound examination (Table 3). Furthermore, plasma MMP-7 and TIMP-1 levels were associated with increased total wall volume as determined by MRI, when modeled in tertiles. Normalized wall index (NWI) was associated with increased levels of MMP-3, MMP-7, and TIMP-1. Although NWI is recognized as an indicator of plaque burden, the measurement itself does not directly involve the lipid core or fibrous cap. The elevated plasma levels of MMPs and TIMP-1 associated with measures of increased carotid wall thickness may reflect their increased production leading to active arterial wall remodeling.
In addition to early atherosclerosis, MMPs have also been associated with plaque progression and development of the so-called vulnerable plaque.10, 22 Although there are discordant reports concerning the effect of calcium on plaque stability, arterial calcification has often been associated with plaque vulnerability.26, 27 Alternatively, arterial calcification may simply reflect overall atherosclerotic burden. In the present study, a significant association was found between plasma MMP-7 level and calcium area, suggesting a possible role for MMP-7 in arterial calcification.
Circulating levels of TIMP-1 were positively associated with presence of lipid core and lipid core area, as measured by MRI. Furthermore, elevated plasma TIMP-1 was associated with increased mean fibrous cap thickness. Taken together, our findings suggest that TIMP-1 may be involved in processes leading to thickening of the carotid artery wall and plaque progression into plaques with large lipid cores. However, the positive association of plasma TIMP-1 levels with mean fibrous cap thickness may point to a possible role for TIMP-1 in plaque stabilization. A number of in vitro studies as well as studies in animals and humans have investigated the possible role of TIMP-1 in the etiology of atherosclerosis. However, to this date, there is only limited data available pointing to a possible causal link between TIMP-1 and atherosclerotic plaque progression. TIMP-1 has been shown to have a mitogenic effect on human aortic smooth muscle cells (SMC) suggesting that it may contribute to SMC proliferation in atherosclerosis.28 In a prospective study of patients undergoing coronary angiography, elevated plasma TIMP-1 was an independent predictor of all-cause mortality and myocardial infarction.29 Furthermore, increased circulating levels of TIMP-1 were associated with the presence of carotid artery lesions in a study of subjects with hyperlipidemia.30 In contrast, we recently reported that plasma TIMP-1 levels were not predictive of incident coronary artery disease in a case-random cohort sample of the ARIC study.31 However, it is plausible that differences in study design and population may be partly responsible for these seemingly contradictory findings. Alternatively, TIMP-1 could be up-regulated in response to a generalized MMP over-expression as a protective mechanism to prevent excessive ECM degradation and plaque destabilization.22
In the present study several factors could explain the larger number of significant associations observed between MMPs and TIMP-1 and variables related to general measures of the carotid wall compared to MRI variables specific to plaque. First, determinations such as wall thickness include the entire study group of 1,678 whereas measures of plaque characteristics are limited to only the 569 participants with a detectable lipid core, leading to differences in statistical power. Secondly, the resolution of the MRI limits the ability to characterize small plaques, restricting the study to the largest plaques found within the thickest arterial walls. Therefore, by truncating the distribution of lipid core volume and cap thickness, possible associations with MMP levels may have gone undetected in our study.
There are several limitations to the current study. First, MMPs/TIMP-1 were measured in the circulation rather than in the arterial wall itself. While it has been demonstrated that a number of different cell types present in atherosclerotic lesions are capable of MMP and TIMP synthesis, it is unlikely that circulating MMP levels solely reflect their release from atherosclerotic plaques. Instead, it is more likely that systemic MMP and TIMP levels reflect their general release from the vasculature, as well as interstitial compartments and other non-vascular sources. However, a number of investigators have reported associations between circulating MMPs and TIMP-1 levels and severity of atherosclerosis.11, 12, 32, 33
Secondly, total MMPs and TIMP-1 were measured, but it is the active forms of the MMPs that are responsible for their enzymatic functions. All of the MMPs studied here are secreted as the pro-enzyme, so this zymogen form is initially present before activation. If the synthesis and secretion of MMPs/TIMP-1 are increased in response to factors that regulate arterial remodeling events, then an increase in the total forms of the MMPs might be an indicator of increased remodeling activity. It would be most informative to measure the active forms in conjunction with the total forms of these proteases. The additional measurement of the active forms of the MMPs and their association with arterial wall characteristics would be desirable in future studies, since such data are lacking.
There are also strengths of the current study. First, the study group was selected from the well-characterized biracial population-based ARIC cohort, and to our knowledge, this is the largest population-based study investigating the relationship of several MMPs and TIMP-1 with quantitative measures of carotid wall components using a gadolinium contrast-enhanced MRI exam. Secondly, a standardized MRI protocol with central reading facility was utilized and extensive quality control data were collected allowing assessment of the reliability of the MRI measures as well as MMPs/TIMP-1 plasma concentrations.
A number of investigators have noted the importance of blood collection procedures especially concerning the measurements of circulating MMPs and TIMP-1.34, 35 There is great discordance in the literature regarding the effects of MMPs and TIMPs, which could be due in part to blood collection procedures. Serum levels of a number of these analytes are influenced by the release of MMPs and TIMP-1 following degranulation of leukocytes and platelets during the ex-vivo blood clotting process in the specimen collection tube. The extent of this elevation varies for the specific analyte measured, but TIMP-1, MMP-1, MMP-8, and MMP-9 have been measured 5–20 times higher in serum than plasma. Since such elevated serum levels result in high background levels, it is generally agreed that plasma, the clinical specimen used in this study, is the preferred source for MMP and TIMP-1 measurement.34 When comparing data from clinical studies related to circulating concentrations of MMPs attention should be given to the method of blood collection utilized.
In summary, our data indicate that circulating levels of MMPs and TIMP-1 are associated with distinct carotid wall characteristics. Additional studies are needed to investigate the relationship between MMPs and atherosclerosis burden.
Supplementary Material
Supplement Material
Supplemental Table I. Race- and gender-specific weighted means (SE), proportions of risk factors and plasma levels of MMPs and TIMP-1 in ARIC Carotid MRI Study participants, adjusted by age*
ACKNOWLEDGEMENTS
a) Acknowledgements–The authors thank the staff and participants of the ARIC study for their important contributions.
b) Sources of funding–The Atherosclerosis Risk in Communities study is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022 with the ARIC carotid MRI examination funded by U01HL075572-01.
Footnotes
Disclosure-The authors have nothing to disclose.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Contributor Information
John W. Gaubatz, Baylor College of Medicine and Methodist Debakey Heart and Vascular Center, Houston, TX.
Christie M. Ballantyne, Baylor College of Medicine and Methodist Debakey Heart and Vascular Center, Houston, TX.
Bruce A. Wasserman, Johns Hopkins Hospital, Baltimore, MD.
Max He, University of North Carolina at Chapel Hill, Chapel Hill, NC.
Lloyd E. Chambless, University of North Carolina at Chapel Hill, Chapel Hill, NC.
Eric Boerwinkle, University of Texas Health Science Center at Houston, Houston, TX.
Ron C. Hoogeveen, Baylor College of Medicine and Methodist Debakey Heart and Vascular Center, Houston, TX.
Reference List
- 1.Ross R. Mechanisms of disease - Atherosclerosis - An inflammatory disease. New England Journal of Medicine. 1999;340:115–126. doi: 10.1056/NEJM199901143400207. [DOI] [PubMed] [Google Scholar]
- 2.Hennekens CH. Increasing burden of cardiovascular disease - Current knowledge and future directions for research on risk factors. Circulation. 1998;97:1095–1102. doi: 10.1161/01.cir.97.11.1095. [DOI] [PubMed] [Google Scholar]
- 3.Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodeling and atherogenesis - The good, the bad, and the ugly. Circulation Research. 2002;90:251–262. [PubMed] [Google Scholar]
- 4.Nagase H, Woessner JF. Matrix metalloproteinases. Journal of Biological Chemistry. 1999;274:21491–21494. doi: 10.1074/jbc.274.31.21491. [DOI] [PubMed] [Google Scholar]
- 5.Keeling WB, Armstrong PA, Stone PA, Bandyk DF, Shames ML. An overview of matrix metalloproteinases in the pathogenesis and treatment of abdominal aortic aneurysms. Vasc.Endovascular Surg. 2005;39:457–464. doi: 10.1177/153857440503900601. [DOI] [PubMed] [Google Scholar]
- 6.Beaudeux JL, Giral P, Bruckert E, Foglietti MJ, Chapman MJ. Matrix metalloproteinases and atherosclerosis. Therapeutic aspects. Annales de Biologie Clinique. 2003;61:147–158. [PubMed] [Google Scholar]
- 7.Rouis M, Adamy C, Duverger N, Lesnik P, Horellou P, Moreau M, Emmanuel F, Caillaud JM, Laplaud PM, Dachet C, Chapman MJ. Adenovirus-mediated overexpression of tissue inhibitor of metalloproteinase-1 reduces atherosclerotic lesions in apolipoprotein E-deficient mice. Circulation. 1999;100:533–540. doi: 10.1161/01.cir.100.5.533. [DOI] [PubMed] [Google Scholar]
- 8.Choudhary S, Higgins CL, Chen IY, Reardon M, Lawrie G, Vick GW, III, Karmonik C, Via DP, Morrisett JD. Quantitation and localization of matrix metalloproteinases and their inhibitors in human carotid endarterectomy tissues. Arteriosclerosis, Thrombosis, and Vascular Biology. 2006;26:2351–2358. doi: 10.1161/01.ATV.0000239461.87113.0b. [DOI] [PubMed] [Google Scholar]
- 9.Galis ZS, Sukhova GK, Lark MW, Libby P. Increased Expression of Matrix Metalloproteinases and Matrix-Degrading Activity in Vulnerable Regions of Human Atherosclerotic Plaques. Journal of Clinical Investigation. 1994;94:2493–2503. doi: 10.1172/JCI117619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kai HK, Ikeda H, Yasukawa H, Kai M, Seki Y, Kuwahara F, Ueno T, Sugi K, Imaizumi T. Peripheral blood levels of matrix metalloproteases-2 and -9 are elevated in patients with acute coronary syndromes. Journal of the American College of Cardiology. 1998;32:368–372. doi: 10.1016/s0735-1097(98)00250-2. [DOI] [PubMed] [Google Scholar]
- 11.Noji Y, Kajinami K, Kawashiri M, Todo Y, Horita T, Nohara A, Higashikata T, Inazu A, Koizumi J, Takegoshi T, Mabuchi H. Circulating matrix metalloproteinases and their inhibitors in premature coronary atherosclerosis. Clinical Chemistry and Laboratory Medicine. 2001;39:380–384. doi: 10.1515/CCLM.2001.060. [DOI] [PubMed] [Google Scholar]
- 12.The Ai. The Atherosclerosis Risk in Communities (ARIC) Study - Design and Objectives. American Journal of Epidemiology. 1989;129:687–702. [PubMed] [Google Scholar]
- 13.White AD, Folsom AR, Chambless LE, Sharret AR, Yang K, Conwill D, Higgins M, Williams OD, Tyroler HA. Community surveillance of coronary heart disease in the atherosclerosis risk in communities (ARIC) study: Methods and initial two years’ experience. Journal of Clinical Epidemiology. 1996;49:223–233. doi: 10.1016/0895-4356(95)00041-0. [DOI] [PubMed] [Google Scholar]
- 14.Wasserman BA, Smith WI, Trout HH, Cannon RO, Balaban RS, Arai AE. Carotid artery atherosclerosis: In vivo morphologic characterization with gadolinium-enhanced double-oblique MR imaging - Initial results. Radiology. 2002;223:566–573. doi: 10.1148/radiol.2232010659. [DOI] [PubMed] [Google Scholar]
- 15.Wagenknecht L, Wasserman B, Chambless L, Coresh J, Folsom A, Mosley T, Ballantyne C, Sharrett R, Boerwinkle E. Correlates of carotid plaque presence and composition as measured by MRI. Circulation: Cardiovascular Imaging. 2009;2:314–322. doi: 10.1161/CIRCIMAGING.108.823922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Saam T, Yuan C, Chu B, Takaya N, Underhill H, Cai J, Tran N, Polissar NL, Neradilek B, Jarvik GP, Isaac C, Garden GA, Maravilla KR, Hashimoto B, Hatsukami TS. Predictors of carotid atherosclerotic plaque progression as measured by noninvasive magnetic resonance imaging. Atherosclerosis. 2007;194:e34–e42. doi: 10.1016/j.atherosclerosis.2006.08.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zhao X, Miller ZE, Yuan C. Atherosclerotic plaque imaging by carotid MRI. Curr Cardiol Rep. 2009;11:70–7. doi: 10.1007/s11886-009-0011-2. [DOI] [PubMed] [Google Scholar]
- 18.Bonnema DD, Webb CS, Pennington WR, Stroud RE, Leonardi AE, Clark LL, McClure CD, Finklea L, Spinale FG, Zile MR. Effects of age on plasma matrix metalloproteinases (MMPs) and tissue inhibitor of metalloproteinases (TIMPs). J.Card Fail. 2007;13:530–540. doi: 10.1016/j.cardfail.2007.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Samnegard A, Silveira A, Lundman P, Boquist S, Odeberg J, Hulthe J, McPheat W, Tornvall P, Bergstrand L, Ericsson CG, Hamsten A, Eriksson P. Serum matrix metalloproteinase-3 concentration is influenced by MMP-3 -1612 5A/6A promoter genotype and associated with myocardial infarction. Journal of Internal Medicine. 2005;258:411–419. doi: 10.1111/j.1365-2796.2005.01561.x. [DOI] [PubMed] [Google Scholar]
- 20.Luan ZX, Chase AJ, Newby AC. Statins inhibit secretion of metalloproteinases-1,-2,-3, and-9 from vascular smooth muscle cells and macrophages. Arteriosclerosis Thrombosis and Vascular Biology. 2003;23:769–775. doi: 10.1161/01.ATV.0000068646.76823.AE. [DOI] [PubMed] [Google Scholar]
- 21.Son JW, Koh KK, Ahn JY, Jin DK, Park GS, Kim DS, Shin EK. Effects of statin on plaque stability and thrombogenicity in hypercholesterolemic patients with coronary artery disease. International Journal of Cardiology. 2003;88:77–82. doi: 10.1016/s0167-5273(02)00368-6. [DOI] [PubMed] [Google Scholar]
- 22.Newby AC. Metalloproteinase expression in monocytes and macrophages and its relationship to atherosclerotic plaque instability. Arterioscler Thromb Vasc Biol. 2008;28:2108–14. doi: 10.1161/ATVBAHA.108.173898. [DOI] [PubMed] [Google Scholar]
- 23.Sluijter JP, Pulskens WP, Schoneveld AH, Velema E, Strijder CF, Moll F, de Vries JP, Verheijen J, Hanemaaijer R, de Kleijn DP, Pasterkamp G. Matrix metalloproteinase 2 is associated with stable and matrix metalloproteinases 8 and 9 with vulnerable carotid atherosclerotic lesions: a study in human endarterectomy specimen pointing to a role for different extracellular matrix metalloproteinase inducer glycosylation forms. Stroke. 2006;37:235–239. doi: 10.1161/01.STR.0000196986.50059.e0. [DOI] [PubMed] [Google Scholar]
- 24.Thrailkill KM, Moreau CS, Cockrell G, Simpson P, Goel R, North P, Fowlkes JL, Bunn RC. Physiological matrix metalloproteinase concentrations in serum during childhood and adolescence, using Luminex Multiplex technology. Clin.Chem.Lab Med. 2005;43:1392–1399. doi: 10.1515/CCLM.2005.238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sternlicht MD, Werb Z. How matrix metalloproteinases regulate cell behavior. Annual Review of Cell and Developmental Biology. 2001;17:463–516. doi: 10.1146/annurev.cellbio.17.1.463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Doherty TM, Detrano RC, Mautner SL, Mautner GC, Shavelle RM. Coronary calcium: The good, the bad, and the uncertain. American Heart Journal. 1999;137:806–814. doi: 10.1016/s0002-8703(99)70403-5. [DOI] [PubMed] [Google Scholar]
- 27.Miralles M, Merino J, Busto M, Perich X, Barranco C, Vidal-Barraquer F. Quantification and characterization of carotid calcium with multi-detector CT-angiography. European Journal of Vascular and Endovascular Surgery. 2006;32:561–567. doi: 10.1016/j.ejvs.2006.02.019. [DOI] [PubMed] [Google Scholar]
- 28.Akahane T, Akahane M, Shah A, Thorgeirsson UP. TIMP-1 stimulates proliferation of human aortic smooth muscle cells and Ras effector pathways. Biochemical and Biophysical Research Communications. 2004;324:440–445. doi: 10.1016/j.bbrc.2004.09.063. [DOI] [PubMed] [Google Scholar]
- 29.Cavusoglu E, Ruwende C, Chopra V, Yanamadala S, Eng C, Clark LT, Pinsky DJ, Marmur JD. Tissue inhibitor of metalloproteinase-1 (TIMP-1) is an independent predictor of all-cause mortality, cardiac mortality, and myocardial infarction. American Heart Journal. 2006;151:1101.e1–1101.e8. doi: 10.1016/j.ahj.2006.02.029. [DOI] [PubMed] [Google Scholar]
- 30.Beaudeux JL, Giral P, Bruckert E, Bernard M, Foglietti MJ, Chapman MJ. Serum matrix metalloproteinase-3 and tissue inhibitor of metalloproteinases-1 as potential markers of carotid atherosclerosis in infraclinical hyperlipidemia. Atherosclerosis. 2003;169:139–146. doi: 10.1016/s0021-9150(03)00149-7. [DOI] [PubMed] [Google Scholar]
- 31.Nambi V, Morrison AC, Hoogeveen RC, Coresh J, Miles S, Rhodes CE, Sharrett AR, Boerwinkle BE, Ballantyne CM. Matrix metalloproteinase-1 and tissue inhibitors do not predict incident coronary artery disease in the atherosclerosis risk in communities (ARIC) study. Tex Heart Inst J. 2008;35:388–94. [PMC free article] [PubMed] [Google Scholar]
- 32.Sapienza P, di Marzo L, Borrelli V, Sterpetti AV, Mingoli A, Cresti S, Cavallaro A. Metalloproteinases and their inhibitors are markers of plaque instability. Surgery. 2005;137:355–363. doi: 10.1016/j.surg.2004.10.011. [DOI] [PubMed] [Google Scholar]
- 33.Altieri P, Brunelli C, Garibaldi S, Nicolino A, Ubaldi S, Spallarossa P, Olivotti L, Rossettin P, Barsotti A, Ghigliotti G. Metalloproteinases 2 and 9 are increased in plasma of patients with heart failure. European Journal of Clinical Investigation. 2003;33:648–656. doi: 10.1046/j.1365-2362.2003.01187.x. [DOI] [PubMed] [Google Scholar]
- 34.Rossignol P, Cambillau M, Bissery A, Mouradian D, Benetos A, Michel JB, Plouin PF, Chatellier G, Jacob MP. Influence of blood sampling procedure on plasma concentrations of matrix metalloproteinases and their tissue inhibitors. Clin.Exp.Pharmacol.Physiol. 2008;35:464–469. doi: 10.1111/j.1440-1681.2008.04897.x. [DOI] [PubMed] [Google Scholar]
- 35.Gerlach RF, Demacq C, Jung K, Tanus-Santos JE. Rapid separation of serum does not avoid artificially higher matrix metalloproteinase (MMP)-9 levels in serum versus plasma. Clinical Biochemistry. 2007;40:119–123. doi: 10.1016/j.clinbiochem.2006.10.007. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplement Material
Supplemental Table I. Race- and gender-specific weighted means (SE), proportions of risk factors and plasma levels of MMPs and TIMP-1 in ARIC Carotid MRI Study participants, adjusted by age*