The Dallas Heart Study (DHS) is a probability-based population sample of 6101 Dallas County residents.15
African Americans and women were oversampled intentionally to achieve a final cohort of approximately 50% African Americans and 50% women. Following an initial in-home visit for collection of survey data, body mass index and measurement of blood pressure, participants between the ages of 30 and 65 were invited to participate in a second visit where they provided in-home fasting blood and urine specimens. Those completing visit two were invited to and to a third visit at UT Southwestern Medical Center, where imaging studies including cardiac magnetic resonance imaging and electron beam computed tomography were performed. Demographics, blood pressure and body composition were similar between subjects completing visits 1 and 2, and laboratory data were similar between those completing visits 2 and 3.15
The present study includes 3294 DHS subjects from visit 2 who underwent measurement of MPO.
Definition of Variables
Hypertension was defined as an average systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or use of anti-hypertensive medication. Hypercholesterolemia was defined as a fasting low density lipoprotein ≥160 mg/dL, a total cholesterol ≥240 mg/dL, or use of a statin medication. Low high density lipoprotein was defined as HDL-C <40 mg/dL in men and <50mg/dL in women. Diabetes was defined as a fasting blood glucose level of ≥126 mg/dL, or a non-fasting blood glucose level of ≥200 mg/dL, or self-reported diabetes with use of any glucose lowering medication. Body mass index was calculated based on measured body weight and height. Race was self-reported.
Personnel who performed atherosclerosis measurements were blinded to all participant data. Coronary artery calcification (CAC) was determined as the average score on two consecutive electron beam computed tomography (EBCT) scans. Agreement between the two scans, as assessed by the intraclass correlation coefficient, was 0.96. Prevalent CAC was defined as >10 Agastaton units, which was a data-derived threshold selected to maximize the signal to noise ratio, as previously described.16
The abdominal aorta was assessed using a 1.5 Tesla whole-body MRI system (Intera, Philips Medical Systems), using a free-breathing, ECG-gated, T2-weighted turbo spin-echo (black-blood) sequence, with 6 transverse slices of the infrarenal abdominal aorta obtained as described previously.17
Aortic plaque (AP) was defined as a hyper-intense signal volume that protruded ≥1 mm from the endoluminal surface of the aortic wall, and was manually contoured in each image.18
Total vascular area (TVA) and total plaque area (TPA) were calculated as the summation of vessel area and plaque area for all 6 slices. Aortic plaque burden (APB) was then calculated by the formula: 100 × (TPA/TVA). Aortic wall thickness (AWT) was calculated by dividing the total vessel wall area by the aortic circumference in each slice, as previously described.19
Mean AWT was then determined by the summation of AWT for each slice divided by number of total slices (n=6). In 70 subjects, an interobserver variability analysis demonstrated an intraclass correlation coefficient between two observers of 0.94 and a mean interobserver difference of 4.2 + 6.6%.20
Measurement of MPO and Other Biomarkers
Venous blood was collected in standard blood collection tubes containing citrate EDTA and samples were maintained at 4°C for ≤4 hours and then centrifuged (1430g for 15 minutes) at 4°C. Plasma was then removed, aliquoted, and frozen at −80°C until assays were performed. MPO was measured from thawed frozen plasma at Alere San Diego, Inc (San Diego, CA) Inc using a sandwich assay on a Luminex 200 reader (Austin, TX) and modified paramagnetic Luminex beads from Radix Biosolutions (Georgetown, TX) with minimum and maximum detection limits of 0.2 ng/mL and 250 ng/mL, respectively. The intra-assay coefficient of variation (CV) was 12% and inter-assay CV was 13%. Personnel who performed the assays were blinded to all clinical data.
The following analytes were measured previously and the methods have been described: high sensitivity C-reactive protein (hsCRP)14
, interleukin-18 (IL18)12
, Tumor necrosis factor-alpha 1 receptor (TNFR1A)21
, matrix metalloproteinase-9 (MMP-9)21
, pulmonary surfactant protein-B (SP-B)21
, monocyte chemoattractant protein (MCP-1)22
, and soluble receptor for advanced glycation end products (sRAGE)23
Participants were divided into quartiles on the basis of MPO levels. Demographic and clinic variables were compared across MPO quartiles using the χ2 trend test for categorical variables and the Jonckheere-Terpstra test for continuous variables. Correlations between selected biomarkers and MPO were evaluated by Spearman rank correlation coefficients. Logistic regression was performed to investigate associations between MPO and prevalent CAC and AP in unadjusted models and models adjusted for traditional risk factors including age, sex, race/ethnicity, body mass index, diabetes, current smoking, hypertension, hypercholesterolemia and low HDL. Linear regression was performed to assess associations between MPO and AWT and APB in unadjusted models and models adjusting for the same traditional risk factors. Sensitivity analyses were also performed with MPO entered into the models as a log-transformed continuous variable. Testing for statistical interaction was performed for MPO × race/ethnicity for all the atherosclerosis phenotypes. Because significant race × MPO interactions were seen, stratified analyses were performed in subgroups defined by race/ethnicity. All models included only subjects with complete data available for covariates and the phenotype of interest. For all statistical testing, 2-sided probability values were reported and a probability value <0.05 was considered statistically significant. No adjustment for multiple testing was done. All analyses were performed using SAS 9.2 (Cary, NC, USA) and all box-plot figures were created using GraphPad PRISM 5.01 (La Jolla, CA, USA).