Coronary Risk Factors and Myocardial Perfusion in Asymptomatic Adults

Transcription

1 Journal of the American College of Cardiology Vol. 47, No. 3, by the American College of Cardiology Foundation ISSN /06/$32.00 Published by Elsevier Inc. doi: /j.jacc Coronary Risk Factors and Endothelial Dysfunction Coronary Risk Factors and Myocardial Perfusion in Asymptomatic Adults The Multi-Ethnic Study of Atherosclerosis (MESA) Lu Wang, MD, PHD,* Michael Jerosch-Herold, PHD, David R. Jacobs, JR, PHD,* Eyal Shahar, MD, MPH,* Aaron R. Folsom, MD, MPH* Minneapolis, Minnesota; Portland, Oregon; and Oslo, Norway OBJECTIVES BACKGROUND METHODS RESULTS CONCLUSIONS The purpose of this study was to determine the cross-sectional relation between myocardial perfusion and coronary heart disease (CHD) risk factors among adults with no clinical CHD. Clinical studies suggest that myocardial perfusion is often abnormal in individuals without CHD but with risk factors. Epidemiologic study in asymptomatic populations is lacking. Two hundred twenty-two men and women, ages 45 to 84 years and free of a CHD diagnosis, in the University of Minnesota field center of the Multi-Ethnic Study of Atherosclerosis (MESA) had myocardial blood flow (MBF) determined using cardiac magnetic resonance imaging at rest and during adenosine-induced hyperemia. Perfusion reserve (PR) was calculated as the ratio of hyperemic to rest MBF. Both resting and hyperemic MBF were lower in men than in women, even after considering age and menopause. Hyperemic MBF was also significantly lower in subjects who were older, and in those with higher blood pressure, higher fasting glucose, and lower low-density lipoprotein cholesterol. After adjusting for age, gender, and race, reduced PR was independently associated with hypertension, higher diastolic blood pressure, and higher total and low-density lipoprotein cholesterol, but was not associated with cigarette smoking, obesity, physical activity, or diabetes. Moreover, hyperemic MBF and PR were correlated strongly and inversely with estimated 10-year CHD risk based on Framingham equations (p for trends: ). Coronary vasoreactivity is reduced in asymptomatic individuals with a greater coronary risk factor burden. Our study results imply that changes in coronary vascular reactivity, in response to risk factors, may be detected in adults without symptomatic CHD. (J Am Coll Cardiol 2006;47:565 72) 2006 by the American College of Cardiology Foundation Myocardial blood flow (MBF) can be quantitatively measured by noninvasive imaging of the transit of intravenously injected contrast agents through the myocardium. The maximal MBF achieved during hyperemia, compared to MBF at rest, reflects the capacity of the coronary vasculature to self-regulate vasomotion. Clinical studies suggest that MBF is often abnormal in individuals without evidence of coronary heart disease (CHD) but with risk factors (1 10), and the coronary flow reserve can be improved with lifestyle modifications or pharmacological intervention (11 13). However, the association between MBF and CHD risk factors has not been investigated in population-based studies of asymptomatic adults. The objective of this study was to determine the crosssectional relation between myocardial perfusion, studied From the *Division of Epidemiology, School of Public Health and the Department of Radiology, School of Medicine, University of Minnesota, Minneapolis, Minnesota; Advanced Imaging Research Center, Oregon Health & Science University, Portland, Oregon; and the Department of Nutrition, University of Oslo, Oslo, Norway. Supported by grant R01 HL and contracts N01-HC through N01-HC (MESA) from the National Heart, Lung, and Blood Institute, National Institutes of Health. Manuscript received May 20, 2005; revised manuscript received September 5, 2005, accepted September 19, with a validated cardiac magnetic resonance imaging (CMR) technique at rest and during adenosine-induced hyperemia, and CHD risk factors in asymptomatic individuals. We hypothesized that perfusion reserve (PR), calculated as the ratio of hyperemic MBF to resting MBF, is inversely associated with levels of CHD risk factors. METHODS Study subjects. Subjects were recruited from participants in the University of Minnesota field center of the Multi- Ethnic Study of Atherosclerosis (MESA), a populationbased prospective cohort study of subclinical cardiovascular disease and its progression (14). All participants in the MESA study reported no previous history of CHD. Each cohort member at the MESA Minnesota Field Center was contacted for a perfusion study, either immediately after the baseline MESA clinic exam or later, by mail. Two hundred thirty-four of the 1,066 cohort members finally consented to be examined for myocardial perfusion, a mean of nine months (range: 0.5 to 33 months) after their first MESA clinic exams in 2000 to Members of this sub-cohort were 45 to 84 years old; 57% were Caucasian (the remainder Hispanic) and 57% were male, which is similar to the entire

2 566 Wang et al. JACC Vol. 47, No. 3, 2006 Coronary Risk Factors and Myocardial Perfusion February 2006: Abbreviations and Acronyms ANOVA analysis of variance AV atrial-ventricular BMI body mass index CHD coronary heart disease CMR cardiac magnetic resonance imaging HDL high-density lipoprotein IV intravenous LDL low-density lipoprotein MBF myocardial blood flow MESA Multi-Ethnic Study of Atherosclerosis PR perfusion reserve RPP rate-pressure product SI signal intensity MESA Minnesota cohort. Subjects who participated in the perfusion study were comparable to those who did not on most standard risk factors, but had a significantly (p 0.05) smaller body mass index (28.7 vs kg/m 2 ) and a lower prevalence of hypertension (21% vs. 33%). MR image acquisition. Participants were asked to abstain from caffeine intake for 12 h before their CMR exam. Cardiac magnetic resonance imaging was performed with a 1.5-T clinical magnetic resonance scanner (Sonata, Siemens Medical Systems, Iselin, New Jersey) at the Fairview- University of Minnesota hospital. The participant was positioned supine with an intravenous (IV) needle inserted into an antecubital vein for injection of MR contrast and infusion of adenosine. A flexible, four-element phased array coil was placed on the participant s chest at the level of the heart; two elements of a spine array coil integrated into the patient table were used as posterior coil elements. Scout magnetic resonance images were acquired first, to determine the orientation of the long and short axis of the left ventricle. Blood pressure, heart rate, and the electrocardiogram were monitored and recorded during the CMR examination. Rate-pressure product (RPP), at rest and hyperemia, was calculated as the product of heart rate (in beats/min) multiplied by systolic blood pressure (in mm Hg) during the scan and divided by 10,000. We conducted T1-weighted imaging with a fast gradient echo sequence for two to three slices in a short-axis orientation (slice thickness: 8 mm) to track the first pass of an injected contrast agent bolus through the right and left ventricle and its recirculation. A Gd-DTPA bolus (Magnevist, Berlex, Wayne, New Jersey) of 0.04 mmol per kg of body weight (total volume mean SD: ml) was injected, starting at the third or fourth heartbeat, with a power injector at a rate of 7 ml/s, followed by a saline flush of 10 ml at the same injection rate. All slices were imaged during each heartbeat, for a total of 50 heart beats. The number of slices was adjusted to two or three to meet the overriding priority of achieving a temporal resolution equivalent to each participant s R-to-R duration. A first perfusion scan was performed during rest, followed by a second scan about 15 min later during maximal vasodilation. Vasodilation was induced by intravenous infusion of adenosine: 0.14 mg/kg/min for 3 min before start of the scan, blocked for approximately 3 s during MR contrast injection, and resumed immediately thereafter. The adenosine infusion was not completely turned off until acquisition of the first 10 to 15 images. Brief, asymptomatic atrial-ventricular (AV) blocks (duration 5 s) were observed in 43 cases, all of which occurred after the contrast bolus injection, and adenosine had already been infused for an initial 3 min. The AV blocks resolved within a few seconds after the adenosine infusion was stopped. The perfusion scans were completed, and hyperemic MBF was determined in all cases where adenosine infusion was prematurely terminated. All MBF measurements were used in statistical analyses. Repeated analyses eliminating those participants who had a brief AV block yielded relatively weaker results. Image analysis and MBF quantification. Endocardial and epicardial contours were manually traced. The myocardium was subdivided into eight transmural sectors of equal circumferential extent along the myocardial centerline. Region-of-interest signal intensity (SI) curves were generated with the MASS CMR image analysis software (Laboratory for Clinical and Experimental Image Processing, Leiden University, Leiden, the Netherlands). Signal intensity curves represent the change of mean SI in each myocardial sector as a function of time. The mean SI before appearance of the contrast agent in the left ventricle was subtracted for baseline correction of each signal curve. In accordance with the central volume principle (15), MBF was estimated from the initial amplitude of the myocardial impulse response by deconvolution analysis of the myocardial SI curves. To perform a model-independent deconvolution of the SI curves, custom-written software was used with an arterial input measured in the center of the left ventricle (16,17). Perfusion reserve was calculated as the ratio of MBF during hyperemia to rest. Since no focal perfusion defect was observed in any participant, all measurements reported in the present study are global averages over the eight myocardial segments and two to three slices. Measurement of risk factors. Risk factors were measured at the MESA clinic using standardized methods. Ever smoking was defined as lifetime consumption of more than 100 cigarettes. Current smoking was defined as smoking cigarettes within the past 30 days. Former smoking was defined as ever but not current smoking. Level of physical activity was computed by multiplying total time of activity per week by activity intensity values. Height and weight were measured, and body mass index (BMI, kg/m 2 ) was assessed. Resting seated blood pressure was measured three times using an automated oscillometric sphygmomanometer, and the average of the last two measurements was used for analysis (called clinic blood pressure ). Hypertension was defined as systolic blood pressure 140 mm Hg, diastolic blood pressure 90 mm Hg, self-reported history of hypertension, or current use of antihypertensive medica-

3 JACC Vol. 47, No. 3, 2006 February 2006: Wang et al. Coronary Risk Factors and Myocardial Perfusion 567 tions. Blood samples were obtained from participants after 8 h of fasting and analyzed at a central laboratory for glucose, total cholesterol, low-density lipoprotein (LDL) cholesterol, and high-density lipoprotein (HDL) cholesterol. Diabetes mellitus was defined as fasting glucose 126 mg/dl, selfreported history of diabetes, or taking diabetes medications. Statistical analysis. Participants were excluded from statistical analysis if they had missing values on any one of the major perfusion measurements (resting MBF, hyperemic MBF or PR, n 5), or they took caffeine within 12 h before CMR scanning (n 7). Analysis was performed using SAS software (SAS Institute, Cary, North Carolina), version 8. The perfusion and hemodynamic measurements were compared across categories of risk factors using t tests or analysis of variance (ANOVA). Pearson correlation coefficients were calculated between continuous risk factors and perfusion measurements. Linear regression analysis was performed to estimate the difference in MBF and PR predicted by a given difference in risk factor, including, if needed, a quadratic term for continuous variables to test the curvilinear trend. Because of potential concerns with the statistical properties of PR measured as a ratio (18,19), we also modeled absolute hyperemic MBF with adjustment for resting MBF as an alternative presentation of PR. Each risk factor was first modeled separately, adjusting only for demographic factors such as age, gender, and race. Afterwards, all the major risk factors, including cigarette smoking habit, BMI, physical activity, hypertension, diabetes, and blood lipids, were entered simultaneously. Finally, to evaluate the joint effect of risk factors on perfusion, 10-year risk of CHD was estimated using the Framingham risk scoring method (20). Perfusion measures were compared across categories of absolute and relative CHD risk using ANOVA. A value of p 0.05 (two-sided) was considered to indicate statistical significance. RESULTS During the rest phase of the perfusion study, the 222 participants had a mean SD heart rate of beats/min, systolic blood pressure of mm Hg, and diastolic blood pressure of mm Hg. Global resting MBF averaged ml/g/min (range: 0.54 to 1.82 ml/g/min) and was correlated positively with resting RPP (r 0.57, p ). Resting RPP was progressively higher with increasing age, BMI, clinic blood pressure, glycemia, and total cholesterol (Table 1). Resting MBF tended to be higher for participants with these same characteristics, except glycemia. Resting MBF was also higher in women and in participants with higher HDL cholesterol. In the participants without AV block, the following hemodynamic parameters changed significantly during hyperemia compared to baseline: diastolic blood pressure decreased by 0.54 mm Hg, systolic blood pressure decreased by 5.2 mm Hg, and heart rate increased by 17 beats/min. In the subgroup of participants with AV block, the hemodynamic changes were milder but still statistically significant, except for systolic blood pressure. Overall, hyperemia raised RPP only slightly (from 0.91 to 1.09, p ), but accompanied by a three-fold increase of MBF to ml/g/min (range: 0.98 to 5.63). The MBF increase was even greater in participants with AV block compared to those with normal response to adenosine, possibly because of the increase in heart rate by 10 to 30 beats after a brief AV block. The correlation between MBF and RPP under hyperemia (r 0.28, p ) was weaker than that at rest. Hyperemic RPP did not differ by most risk factors, except for gender and BMI (Table 1). Women continued to have significantly higher MBF than men during hyperemia, whereas subjects who were older, had higher clinic blood pressure and blood glucose, and had lower HDL cholesterol, had significantly lower hyperemic MBF. Perfusion reserve ranged from 1.18 to 5.24 (mean SD: ). Perfusion reserve was significantly lower among male participants, those at older ages, and those with hypertension and higher levels of fasting glucose, total cholesterol, and LDL cholesterol (Table 1). Age demonstrated the strongest and most significant correlation with PR (r 0.36, p ), followed by clinic systolic blood pressure (r 0.24, p ), diastolic blood pressure (r 0.21, p 0.002), fasting glucose (r 0.16, p 0.02), total cholesterol (r 0.14, p 0.04), and LDL cholesterol (r 0.12, p 0.07). Other risk factors did not show significant correlations with PR. After adjusting for age, gender, and race, participants who had one unit higher diastolic blood pressure, total cholesterol, and LDL cholesterol had lower hyperemic MBF by , , and ml/g/min, respectively, and lower PR by 0.013, , and , respectively (Table 2). Subjects with prevalent hypertension had 0.30 lower PR than did those without hypertension. In the multivariate adjusted model, the inverse relation of PR with hypertension and blood cholesterol remained statistically significant even after adjusting for all the other traditional risk factors simultaneously (Table 3). Stratified analysis revealed that the associations were relatively stronger in the middle-aged participants (45 to 64 years) compared to older participants ( 65 years), but did not differ by gender or medication use. The mean SD predicted 10-year absolute risk of total CHD, estimated from the Framingham prediction equation (20), was 11 8% in this cohort. The predicted absolute CHD risk (Fig. 1) was inversely associated with hyperemic MBF (p for trend: ) and PR (p for trend: ). The prevalence of reduced PR ( 2.5) was progressively higher (16%, 20%, 36%, and 53%, p for trend ) across increasing quartiles of predicted absolute CHD risk. When the excessive risk conferred by age and gender alone was considered, the predicted relative CHD risk had an inverse relation with hyperemic MBF and PR, similar to that for absolute risk (data not shown).

4 568 Wang et al. JACC Vol. 47, No. 3, 2006 Coronary Risk Factors and Myocardial Perfusion February 2006: Table 1. Unadjusted Mean Perfusion Measurements and Hemodynamic Parameters Across Categories of Participant Characteristics Risk Factors N At Rest RPP (beats/min mm Hg) MBF (ml/g/min) During Hyperemia RPP (beats/min mm Hg) MBF (ml/g/min) Perfusion Reserve Age (yrs) p for overall difference Gender Women Men p for t test Race Caucasian Hispanic p for t test Cigarette smoking Never Former Current p for overall difference Body mass index (kg/m 2 ) p for overall difference Physical activity (min/week MET) 1st tertile: 3, nd tertile: 3,500 6, rd tertile: 6, p for overall difference Hypertension No Yes p for t test Systolic blood pressure (mm Hg) p for overall difference Diastolic blood pressure (mm Hg) p for overall difference Diabetes No Yes p for t test Fasting glucose (mg/dl) p for overall difference Total cholesterol (mg/dl) p for overall difference LDL cholesterol (mg/dl) p for overall difference HDL cholesterol (mg/dl) p for overall difference HDL high-density lipoprotein; LDL low-density lipoprotein; MBF myocardial blood flow; RPP rate-pressure product.

5 JACC Vol. 47, No. 3, 2006 February 2006: Wang et al. Coronary Risk Factors and Myocardial Perfusion 569 DISCUSSION Our study shows that the coronary vasodilatory response in terms of hyperemic MBF and PR correlate inversely with CHD risk factor burden, as assessed by Framingham prediction equations, in asymptomatic middle-aged and older adults. The associations vary and tend to be weak with respect to individual risk factors. To our knowledge, this is the first population-based study of MBF under rest and stress conditions in asymptomatic individuals. Myocardial perfusion imaging examination was originally designed to assess the hemodynamic response of obstructive epicardial stenosis. Further research found that impairment of coronary vasoreactivity may appear even before the development of obstructive lesions and therefore might serve as an early marker of subclinical atherosclerosis. Previous non-invasive investigations of MBF in relation to CHD risk factors have been mostly conducted in clinical Table 2. Results for Age, Gender, and Race Adjusted Linear Regression Models of Myocardial Perfusion Measurements (PR and Hyperemic MBF, Respectively) With Coronary Heart Disease Risk Factors as Predictors Hyperemic Blood Flow (ml/min/g) Perfusion Reserve Risk Factors * p * p Age (10.3 yrs), unadjusted Men vs. women, age-adjusted Hispanic vs. Caucasian, age-adjusted Current smoker vs never smoker Former smoker vs never smoker Overweight vs normal weight Obese vs. normal weight Moderate vs. low physical activity High vs. low physical activity Hypertension vs. no Systolic blood pressure (18.7 mm Hg) Diastolic blood pressure (10.1 mm Hg) Diabetes vs. no Fasting glucose (24.0 mg/dl) Total cholesterol (33.7 mg/dl) LDL cholesterol (28.4 mg/dl) HDL cholesterol (13.5 mg/dl) *Regression coefficients were estimated by using a separate model for each specific risk factor, adjusting for age, gender, and race, unless otherwise stated. Regression coefficients for hyperemic blood flow also adjusted for resting blood flow. Value in parentheses is the standard deviation of the continuous risk factor. PR perfusion reserve; other abbreviations as in Table 1. Table 3. Results for Multivariate Adjusted Linear Regression Models of Myocardial Perfusion Measurements (PR and Hyperemic MBF, Respectively) With Coronary Heart Disease Risk Factors as Predictors Hyperemic Blood Flow (ml/min/g) Perfusion Reserve Risk Factors * p * p Current smoker vs never smoker Former smoker vs never smoker Overweight vs normal weight Obesity vs normal weight Moderate vs. low physical activity High vs. low physical activity Hypertension vs. no Diabetes vs. no Total cholesterol (33.7 mg/dl) HDL cholesterol (13.5 mg/dl) *Regression coefficients were estimated adjusting for all the other variables listed in the table as well as age, gender, and race. Regression coefficients for hyperemic blood flow also adjusted for resting blood flow. Value in parentheses is the standard deviation of the continuous risk factor. Abbreviations as in Tables 1 and 2. case-control studies. In agreement with prior work by others (21 25), we observed that resting MBF increases with age whereas hyperemic MBF and PR decline with age; MBF is higher in women than in men, both at rest and during hyperemia, even after correction for RPP and adjustment for age. The significantly higher MBF and PR found in our female participants without obstructive CHD remained even when menopause status and hormone use were considered, which suggests that the gender difference may not be completely explained by a direct effect of estrogen. We also corroborated other studies (26 28) showing that PR induced by endothelium-independent vasodilators was similar in smokers and non-smokers. Myocardial blood flow abnormalities were seen in our asymptomatic subjects with hypertension, which was consistent with earlier findings of perfusion defects in hypertensive patients with no significant CHD (1,2). Besides the increased resting MBF due to the higher myocardial oxygen consumption (29), hypertension also increases total coronary vascular resistance (30), which further reduces PR. Reductions in stress MBF and PR have been previously reported in both type 1 (3 5) and type 2 (4,6,7) diabetes patients in the absence of clinical CHD, attributable to hyperglycemia, hyperinsulinemia and insulin resistance, or a pronounced accumulation of other risk factors. In our study, hyperemic MBF and PR were relatively lower in diabetic than in non-diabetic participants, and fasting blood glucose was inversely correlated with PR. However, these associations disappeared when adjusting for

6 570 Wang et al. JACC Vol. 47, No. 3, 2006 Coronary Risk Factors and Myocardial Perfusion February 2006: Figure 1. Myocardial perfusion measurements examined by magnetic resonance imaging among asymptomatic subjects according to quartiles of predicted absolute 10-year risk of coronary heart disease (CHD): 1% to 4.8% (median: 3%), 4.9% to 7.9% (median: 6%), 8% to 14.9% (median: 11%), 15% to 46% (median: 20%), estimated using Framingham prediction equations with total cholesterol categories (20). The p values for trend are p for rest MBF, and p for hyperemic MBF and PR. The error bar represented 95% confidence interval (CI). other risk factors. One of the more persistent associations we found is for hypercholesterolemia. Compared with normocholesterolemic subjects, in subjects with high levels of total cholesterol or LDL cholesterol, elevated resting MBF combined with decreased hyperemic MBF resulted in a significantly lower PR. Diminished maximal MBF and PR have been seen in patients with familial hypercholesterolemia but without overt atherosclerosis (8 10). Nonsignificant atherosclerotic lesions caused by hypercholesterolemia may increase vessel wall stiffness and weaken the smooth muscle layer (31). Hypercholesterolemia also causes endothelial dysfunction in large conduit vessels, as well as in resistance vessels (32). High HDL cholesterol did not seem to ameliorate unfavorable PR. Compared to previous clinical studies of highly selected patients, the associations of myocardial perfusion with individual risk factors in our study were generally weak. This is not surprising, considering that our study subjects were from a population with, on average, moderate risk factor levels. When the overall risk burden was assessed using the Framingham prediction equations, risk factor burden was strongly and inversely associated with hyperemic MBF and PR. Moreover, in the current study, the inverse correlations of hyperemic MBF and PR with systolic blood pressure and fasting glucose level were substantially attenuated and no longer statistically significant after age adjustment. Such strong confounding by age was not seen in most other studies, probably because those studies usually focused on the association between MBF and single risk factors, such as smoking, hyperlipidemia, and diabetes mellitus, and compared patients with age- and gender-matched control groups. Our study population had a broad age range, and age demonstrated the strongest risk factor correlation with PR; therefore, the remaining effect of other individual risk factors, independent of age, was relatively small. Finally, few studies have addressed MBF in association with obesity and physical activity. In our study, greater physical activity or smaller BMI did not relate to higher age-adjusted hyperemic MBF and PR. Perfusion reserve, calculated as the ratio of hyperemic MBF divided by resting MBF, could be reduced by either a decrease in the numerator or an elevation in the denominator, or both. In the scenario of CHD, when the coronary resistance vessels get sparser, smaller, or less responsive to vasodilatory stimuli, the maximal MBF will be abnormally low while resting MBF may remain normal (18). High resting flow is most likely caused by conditions requiring increased cardiac work and consequently higher myocardial demand for energy, when compensatory mechanisms, e.g., dilation of the distal vascular bed, become involved even at rest. In such cases, the remaining available flow reserve might decrease even if the maximal flow attainable does not diminish. For most risk factors investigated in the present study, the inverse relation with PR resulted primarily from an inverse relation with hyperemic MBF, combined with a trend toward a positive relation with resting MBF. Modeling PR alternatively as hyperemic MBF with adjustment for resting MBF yielded similar results (Table 2). The CMR first-pass technique combined with the model-independent deconvolution method used in our study to quantify MBF was previously validated in experimental animal studies by comparison to MBF measurements with radio-isotope labeled microspheres (16), which is regarded as the reference method for quantification of MBF, although it is not applicable to human studies. The linear correlation of CMR-estimated MBF against measurements from radio-isotope labeled microspheres was excellent (r , slope: 0.96, intercept: 0.06). The reliability of MBF quantification in the current study was generally satisfactory. The coefficient of variation of repeated measurements in each myocardial sector, performed one year apart, averaged % (n 13 subjects) for resting MBF and % (n 8 subjects) for hyperemic MBF. The blood flow response to adenosine is predominantly elicited by relaxation of coronary arteriolar smooth muscle cells. An endothelium-mediated dilation of the proximal vessels, induced by the increased blood flow, could further enhance the total vasodilatory response. Considering atherogenesis to have multiple stages offers another explanation for the dose-response relation we observed between risk factor levels and PR. In some individuals with just mild to moderate CHD risk factor elevations, selective dysfunction of endothelium alone may or may not produce an abnormal vasodilation to adenosine. Among subjects with a more advanced risk profile, a more extensive disorder of endothelial cells and smooth muscle cells may further impair PR. Several issues should be considered as potential limitations of the current study. First of all, in this cross-sectional

7 JACC Vol. 47, No. 3, 2006 February 2006: Wang et al. Coronary Risk Factors and Myocardial Perfusion 571 study, we can only assess the correlation of risk factor levels with concurrent PR. The causal effect of a risk factor on the development of vasoreactivity impairment cannot be established. In addition, the prognostic value of risk factor assessment versus a measurement of myocardial perfusion cannot be compared at this point because of the lack of follow-up data. Secondly, because these asymptomatic participants did not undergo coronary angiography, some may have had obstructive atherosclerotic lesions that were not detected clinically. Coronary calcification was measured by computed tomography in this cohort, but the coronary calcium burden cannot reliably predict stenosis severity or the hemodynamic significance of a calcified lesion. For these reasons, it is difficult to infer a direct effect of CHD risk factors on vascular function, independent of coronary stenosis burden. Thirdly, the majority of subjects in this asymptomatic population had a low to medium predicted 10-year CHD risk. Few participants had extraordinarily adverse risk characteristics. Although our sample size was moderate, we may not have had sufficient statistical power to detect some weak but real associations. One also may not be able to directly generalize our findings from the subgroup of MESA volunteers who participated in the perfusion study, to the whole MESA population due to differences in BMI and the prevalence of hypertension. Finally, we cannot rule out the possibility that the hyperemic MBF response was attenuated by premature stopping of adenosine in participants having an AV block, although this concern was not borne out by a comparison of MBF measured in subjects with versus those without AV block. The high proportion of patients with AV block is probably secondary to use of the same IV line for adenosine and MR contrast. Use of two IV lines could have avoided this complication. Previous studies report that, during intracoronary administration of adenosine in patients, a plateau phase, when the coronary flow velocity remains within 10% of its peak value, lasts approximately 6 s after stopping the adenosine infusion (33,34). Therefore, we expect that when IV adenosine is not resumed after the 3-min initial infusion because of a transient AV block, MBF will stay close to its peak for the initial myocardial contrast-enhancement, which is sufficiently long to estimate hyperemic MBF (35). In conclusion, our study shows that coronary vasoreactivity is reduced in asymptomatic adults with a greater coronary risk factor burden. With respect to individual risk factors, age is correlated inversely with the vasodilatory response. High diastolic blood pressure, total cholesterol, and LDL cholesterol are associated with lower flow response independent of age. These findings corroborate that functional changes in coronary vasculature may occur in asymptomatic adults with risk factors, which supports reduced vasoreactivity as an indicator of CHD before its clinical manifestation. Our investigations improve the understanding of the physiology and pathophysiology role of myocardial perfusion in the process of coronary atherosclerosis. Findings from this study also provide important baseline information for future research on the clinical use of perfusion studies, as well as on the subclinical CHD evaluation. Acknowledgments The authors thank Christine Dwight and Esther Ruiz from the University of Minnesota MESA clinic, for their efforts in recruiting subjects for this study, and Julie Dicken, RN, research nurse in the Division of Cardiology, for her assistance during the MRI perfusion studies. The authors also thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. Reprint requests and correspondence: Dr. Michael Jerosch- Herold, Advanced Imaging Research Center, MS L452, 3181 SW Sam Jackson Park Road, Portland, Oregon jeroschh@ohsu.edu. REFERENCES 1. Gimelli A, Schneider-Eicke J, Neglia D, et al. Homogeneously reduced versus regionally impaired myocardial blood flow in hypertensive patients: two different patterns of myocardial perfusion associated with degree of hypertrophy. J Am Coll Cardiol 1998;31: Parodi O, Neglia D, Sambuceti G, Marabotti C, Palombo C, Donato L. Regional myocardial blood flow and coronary reserve in hypertensive patients. The effect of therapy Drugs 1992;44 Suppl 1: Meyer C, Schwaiger M. Myocardial blood flow and glucose metabolism in diabetes mellitus. Am J Cardiol 1997;80:94A 101A. 4. Di Carli MF, Janisse J, Grunberger G, Ager J. Role of chronic hyperglycemia in the pathogenesis of coronary microvascular dysfunction in diabetes. J Am Coll Cardiol 2003;41: Pitkanen OP, Nuutila P, Raitakari OT, et al. Coronary flow reserve is reduced in young men with IDDM. Diabetes 1998;47: Yokoyama I, Ohtake T, Momomura S, et al. Hyperglycemia rather than insulin resistance is related to reduced coronary flow reserve in NIDDM. Diabetes 1998;47: Yokoyama I, Momomura S, Ohtake T, et al. Reduced myocardial flow reserve in non insulin-dependent diabetes mellitus. J Am Coll Cardiol 1997;30: Yokoyama I, Ohtake T, Momomura S, Nishikawa J, Sasaki Y, Omata M. Reduced coronary flow reserve in hypercholesterolemic patients without overt coronary stenosis. Circulation 1996;94: Pitkanen OP, Raitakari OT, Niinikoski H, et al. Coronary flow reserve is impaired in young men with familial hypercholesterolemia. J Am Coll Cardiol 1996;28: Yokoyama I, Murakami T, Ohtake T, et al. Reduced coronary flow reserve in familial hypercholesterolemia. J Nucl Med 1996;37: Czernin J, Barnard RJ, Sun KT, et al. Effect of short-term cardiovascular conditioning and low-fat diet on myocardial blood flow and flow reserve. Circulation 1995;92: Mellwig KP, Baller D, Gleichmann U, et al. Improvement of coronary vasodilatation capacity through single LDL apheresis. Atherosclerosis 1998;139: Baller D, Gleichmann U, Notohamiprodjo G, et al. [Improved coronary vasodilator capacity by drug lipid lowering therapy in patients in the early stage of coronary atherosclerosis with reduced coronary reserves and moderate LDL hypercholesteremia]. Z Kardiol 1998;87 Suppl 2: Bild DE, Bluemke DA, Burke GL, et al. Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol 2002;156: Zierler KL. Equations for measuring blood flow by external monitoring of radioisotopes. Circ Res 1965;16: Jerosch-Herold M, Swingen C, Seethamraju RT. Myocardial blood flow quantification with MRI by model-independent deconvolution. Med Phys 2002;29:

8 572 Wang et al. JACC Vol. 47, No. 3, 2006 Coronary Risk Factors and Myocardial Perfusion February 2006: Jerosch-Herold M, Seethamraju RT, Swingen CM, Wilke NM, Stillman AE. Analysis of myocardial perfusion MRI. J Magn Reson Imaging 2004;19: Hoffman JI. Maximal coronary flow and the concept of coronary vascular reserve. Circulation 1984;70: Klocke FJ, Ellis AK, Canty JM, Jr. Interpretation of changes in coronary flow that accompany pharmacologic interventions. Circulation 1987;75:V Wilson PW, D Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97: Czernin J, Muller P, Chan S, et al. Influence of age and hemodynamics on myocardial blood flow and flow reserve. Circulation 1993;88: Uren NG, Camici PG, Melin JA, et al. Effect of aging on myocardial perfusion reserve. J Nucl Med 1995;36: Senneff MJ, Geltman EM, Bergmann SR. Noninvasive delineation of the effects of moderate aging on myocardial perfusion. J Nucl Med 1991;32: Rosen SD, Uren NG, Kaski JC, Tousoulis D, Davies GJ, Camici PG. Coronary vasodilator reserve, pain perception, and sex in patients with syndrome X. Circulation 1994;90: Duvernoy CS, Meyer C, Seifert-Klauss V, et al. Gender differences in myocardial blood flow dynamics: lipid profile and hemodynamic effects. J Am Coll Cardiol 1999;33: Meeder JG, Blanksma PK, van der Wall EE, et al. Long-term cigarette smoking is associated with increased myocardial perfusion heterogeneity assessed by positron emission tomography. Eur J Nucl Med 1996;23: Campisi R, Czernin J, Schoder H, et al. Effects of long-term smoking on myocardial blood flow, coronary vasomotion, and vasodilator capacity. Circulation 1998;98: Iwado Y, Yoshinaga K, Furuyama H, et al. Decreased endotheliumdependent coronary vasomotion in healthy young smokers. Eur J Nucl Med Mol Imaging 2002;29: Weber KT, Janicki JS. The metabolic demand and oxygen supply of the heart: physiologic and clinical considerations. Am J Cardiol 1979;44: Schwartzkopff B, Motz W, Frenzel H, Vogt M, Knauer S, Strauer BE. Structural and functional alterations of the intramyocardial coronary arterioles in patients with arterial hypertension. Circulation 1993;88: Lee RT, Kamm RD. Vascular mechanics for the cardiologist. J Am Coll Cardiol 1994;23: Drexler H, Zeiher AM. Endothelial function in human coronary arteries in vivo. Focus on hypercholesterolemia Hypertension 1991;18 Suppl:II Wilson RF, Wyche K, Christensen BV, Zimmer S, Laxson DD. Effects of adenosine on human coronary arterial circulation. Circulation 1990;82: De Bruyne B, Pijls NH, Barbato E, et al. Intracoronary and intravenous adenosine 5=-triphosphate, adenosine, papaverine, and contrast medium to assess fractional flow reserve in humans. Circulation 2003;107: Jerosch-Herold M, Hu X, Murthy NS, Rickers C, Stillman AE. Magnetic resonance imaging of myocardial contrast enhancement with MS-325 and its relation to myocardial blood flow and the perfusion reserve. J Magn Reson Imaging 2003;18: