Author + information
- Received August 11, 2011
- Revision received December 8, 2011
- Accepted December 15, 2011
- Published online August 1, 2012.
- Sammy Elmariah, MD, MPH⁎,†,⁎ (, )
- Joseph A.C. Delaney, PhD‡,
- David A. Bluemke, MD, PhD§,
- Matthew J. Budoff, MD∥,
- Kevin D. O‘Brien, MD¶,
- Valentin Fuster, MD, PhD†,#,
- Richard A. Kronmal, PhD⁎⁎ and
- Jonathan L. Halperin, MD†
- ↵⁎Reprint requests and correspondence:
Dr. Sammy Elmariah, Cardiology Division, Massachusetts General Hospital, Harvard Medical School, 55 Fruit Street, GRB 800, Boston, Massachusetts 02114
Objectives The aim of this study was to evaluate the relationship between percentage of predicted left ventricular mass (%PredLVM) and valve calcification in the MESA (Multi-Ethnic Study of Atherosclerosis) study.
Background Cardiac valve calcification has been associated with left ventricular hypertrophy (LVH), which portends cardiovascular events. However, this relationship and its mediators are poorly understood.
Methods The MESA study is a longitudinal cohort study of men and women 45 to 84 years of age without clinical cardiovascular disease in whom serial cardiac magnetic resonance and computed tomography imaging were performed. The relationships between baseline %PredLVM and the prevalence, severity, and incidence of aortic valve (AVC) and mitral annulus calcification (MAC) were determined by regression modeling.
Results Prevalent AVC was observed in 630, and MAC was observed in 442 of 5,042 subjects (median 55.9 and 71.1 Agatston units, respectively). After adjustment for age, sex, body mass index (BMI), ethnicity, socioeconomic status, physical activity, diabetes, cholesterol levels, blood pressure, smoking, kidney function, serum lipids, and antihypertensive and statin medications, %PredLVM was associated with prevalent AVC (odds ratio [OR]: 1.18/SD increase in %PredLVM [95% confidence interval (CI): 1.08 to 1.30]; p = 0.0004) and MAC (OR: 1.18 [95% CI: 1.06 to 1.32]; p = 0.002). Similarly, %PredLVM was associated with increased severity of prevalent AVC (risk difference = 0.26 [95% CI: 0.15 to 0.38]; p < 0.0001) and MAC (risk difference = 0.20 [95% CI: 0.03 to 0.37]; p = 0.02). During follow-up (mean 2.4 ± 0.9 years), 153 subjects (4%) developed AVC, and 198 (5%) developed MAC. The %PredLVM was associated with incident AVC (OR: 1.24 [95% CI: 1.04 to 1.47]; p = 0.02) and MAC (OR: 1.18 [95% CI: 1.01 to 1.40]; p = 0.04). Further adjustment for inflammatory markers and coronary artery calcification did not attenuate these associations. Specifically, concentric LVH most strongly predicted incident valve calcification.
Conclusions Within the MESA cohort, LVH was associated with prevalence, severity, and incidence of valve calcification independent of hypertension and other identified confounders.
Calcification of the aortic and mitral valves is a progressive disease similar to atherosclerosis (1–4) that is associated with adverse cardiovascular outcomes (5–7). Even without hemodynamically significant valve obstruction, calcific aortic and mitral valve disease have been associated with dramatic increases in the risk of myocardial infarction, stroke, and cardiovascular and all-cause mortality (5–7). Valve calcification might be a marker of atherosclerosis, but discordance between coronary disease and calcific valve disease suggests that alternative mechanisms such as inflammation, neurohormonal activation, endothelial dysfunction, or other genetic factors also might play a role (1–4).
Aortic stenosis (AS) causes compensatory left ventricular hypertrophy (LVH); however, 2 small cross-sectional analyses suggest an association between aortic valve disease and LVH in the absence of significant valve obstruction (3,8). Although LVH in this setting might be a consequence of low levels of outflow obstruction, similar associations between LVH and prevalent mitral annulus calcification (MAC) suggest that alternate processes might lead to both valvular calcification and left ventricular (LV) remodeling (9,10). The longitudinal MESA (Multi-Ethnic Study of Atherosclerosis) study provides a unique opportunity to explore relationships between LVH and calcific valve disease.
Study population and data collection
The MESA study is a prospective cohort study of 6,814 men and women 45 to 84 years of age recruited from 6 U.S. communities designed to evaluate risk factors for cardiovascular disease. At initial enrollment, subjects had no clinical evidence of cardiovascular disease. Participants attended study visits every 18 to 24 months. A detailed description of the study design has been published (11). This analysis was limited to the 5,042 subjects who underwent cardiac magnetic resonance and computed tomography (CT).
Measurement of cardiovascular calcification
Aortic valve and mitral annulus calcification were assessed by electron-beam CT at 3 centers and by multi-detector row helical CT at 3 centers. All studies were interpreted centrally (Los Angeles Biomedical Research Institute at Harbor-UCLA, Torrance, California). Calcification was quantified by Agatston scoring (12). Prevalent cardiovascular calcification was defined as a score >0 Agatston units (AU). Details of the image acquisition and interpretation protocols, quality control measures, and interobserver reliability characteristics have been reported previously (13). Follow-up cardiac CT scans with assessment of aortic valve calcification (AVC) and MAC were performed 2 to 3 years after the initial scan.
Determination of LV mass
Magnetic resonance imaging was performed with 1.5-T magnets with 4-element phased-array surface coils, electrocardiographic gating, and blood pressure monitoring. The LV mass was quantified as previously described (14). With an allometric approach, regression models for body size were derived from a sample of 1,746 MESA participants without obesity, hypertension, antihypertensive medication, diabetes, impaired fasting glucose, or hypoglycemic medication, and a multiplicative estimate was derived from the regression of log(LV mass) on log(height), log(weight), and sex. The LV mass was adjusted for body size by dividing LV mass by the predicted LV mass on the basis of height, weight, and sex as: 100 · LV mass/(a · height0.54 · weight0.61), where a = 6.82 for women and 8.25 for men with mass in grams, height in meters, and weight in kilograms (15). Similarly, LV end-diastolic volume was adjusted for body size and sex by dividing by the predicted LV volume as: 100 · LV end-diastolic volume/(b · height1.25 · weight0.43), where b = 10.0 for women and 10.5 for men. The resultant percentage of predicted LV mass (%PredLVM) and percentage of predicted LV end diastolic volume (%LVVol) were used for all analyses.
Historical data were collected with a combination of self- and interviewer-administered questionnaires. Smoking status was defined as current, former, or never, with current smoking defined as smoking a cigarette in the last 30 days. Diabetes was defined as a fasting glucose ≥126 mg/dl or hypoglycemic medication use. Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, a reported history of hypertension, or antihypertensive therapy. Estimated glomerular filtration rate (eGFR) was calculated with the Modification of Diet in Renal Disease equation. Physical activity was assessed with the MESA Typical Week Physical Activity Survey and quantified for this analysis as minutes of vigorous activity/week multiplied by metabolic equivalent level (16).
Differences in patient characteristics across %PredLVM strata were evaluated with analysis of variance. Chi-square analyses were used for categorical variables. To approximate a normal distribution, calcium scores were log-transformed for use in analyses of severity. Logistic regression was used to derive odds ratios (ORs) for the relationships between baseline %PredLVM and the prevalence and incidence of AVC and MAC. Linear regression models with log-transformed valve calcium scores were used to assess relationships between %PredLVM and severity of AVC and MAC in subjects with prevalent disease. Statistical analyses were performed with SAS (version 9.2, SAS Institute, Inc., Cary, North Carolina), with significance accepted at p < 0.05.
A total of 5,042 subjects underwent both cardiac CT and magnetic resonance imaging scans, with a mean age of 62 ± 10 years; 46% were men (Table 1). The mean end-diastolic LV mass was 145 ± 40 g. The mean %PredLVM, which adjusts for body size and sex, was 104 ± 19%, indicating the LV mass of the population was 4% greater than the reference cohort of participants without obesity, hypertension, antihypertensive medication, diabetes, impaired fasting glucose or hypoglycemic medication. Quartiles of %PredLVM were defined as: quartile 1: <91.5%, quartile 2: 91.5% to 102.2%, quartile 3: 102.2% to 114.6%, quartile 4 ≥114.6% (Table 1). The proportion of subjects with diabetes mellitus, hypertension, and current smoking status increased by %PredLVM quartile. Significant ethnic variation in LV mass were observed with greater %PredLVM observed in Blacks and Hispanics (Table 1).
LV mass and prevalent valve calcium
At the baseline evaluation, cardiac CT identified AVC in 630 subjects (13%) and MAC in 442 (9%). Both AVC and MAC were observed in 183 (4%) subjects. Stratification by %PredLVM quartile demonstrated increasing prevalence of valve calcification with increasing %PredLVM, with AVC observed in 144 subjects (11%) in the lowest and 200 (15%) in the highest %PredLVM quartiles (p = 0.0004) and MAC observed in 97 subjects (8%) in the lowest %PredLVM quartile and 134 (11%) in the highest quartile (p = 0.049) (Fig. 1A). This relationship was consistent across ethnic groups within MESA for both AVC and MAC (Figs. 1B and 1C). After adjusting by multivariable analysis for age, sex, BMI, ethnicity, study site, socioeconomic status, physical activity, diabetes, history of hypercholesterolemia, hypertension, smoking, eGFR, total serum cholesterol and low-density lipoprotein-cholesterol levels, systolic and diastolic blood pressures, and use of antihypertensive agents and statins, %PredLVM was associated with prevalent AVC (OR: 1.18/SD increase in %PredLVM [95% CI: 1.08 to 1.30]; p = 0.0004) and MAC (OR: 1.18 [95% CI: 1.06 to 1.32]; p = 0.002).
LV mass and the severity of valve calcium
Among those subjects with valve calcium at baseline, the median AVC score was 55.9 (interquartile range: 21.0 to 149.1) AU, and the median MAC score was 71.1 (22.4 to 290.2) AU. The severity of both AVC and MAC, defined as log (Agatston score), increased with increasing quartile of %PredLVM (AVC, mean [SD]: 3.7 [1.3], 4.0 [1.3], 4.1 [1.5], 4.3 [1.7], respectively; p = 0.01; MAC: 4.3 [1.9], 4.2 [1.8], 4.5 [1.9], 4.8 [1.8], respectively; p = 0.055). Multivariable regression analyses adjusting for the aforementioned variables demonstrated a robust relationship between %PredLVM and the severity of AVC (risk difference = 0.26/SD increase in %PredLVM [95% CI: 0.15 to 0.38]; p < 0.0001) and MAC (risk difference = 0.20 [95% CI: 0.03 to 0.37]; p = 0.02).
LV mass and incident valve calcium
Over a mean follow-up of 2.4 ± 0.9 years, 153 (4%) of 3,902 subjects without AVC at baseline developed AVC, an annualized incidence rate of 1.6%. Of the 4,072 subjects without MAC at the first evaluation, 198 (5%) developed MAC during follow-up, an incidence rate of 2.0%/year. We found a relationship between incident AVC and baseline %PredLVM (incidence rate 1.2%/year in the lowest quartile and 1.9%/year in the highest quartile; p = 0.05) (Fig. 2). Incident MAC was significantly greater among subjects in the highest %PredLVM quartile than in the lowest quartile (2.9%/year vs. 1.9%/year; p = 0.03).
Because of the racial variations in LV mass and valve calcification and the interaction of age with risk factors for valve calcification (17–19), we tested for interactions between race and age with %PredLVM for incident valve calcification. Neither race nor age demonstrated significant interactions with %PredLVM for AVC (race: Chinese p = 0.43, Black p = 0.98, Hispanic p = 0.20, each vs. Caucasian ethnicity; age: p = 0.63) or MAC (race: Chinese p = 0.97, Black p = 0.41, Hispanic p = 0.98, each vs. Caucasian ethnicity; age: p = 0.20). After adjustment for age, sex, ethnicity, and study site, %PredLVM was significantly associated with incident AVC (OR: 1.20/SD increase in %PredLVM [95% CI: 1.02 to 1.40]; p = 0.03) and incident MAC (OR: 1.16 [95% CI: 1.00 to 1.35]; p = 0.049). Sequential multivariable models were constructed in an effort to identify mediators of this relationship (Table 2). Additional adjustments for cardiovascular risk factors including BMI, diabetes, hypercholesterolemia, hypertension, smoking, eGFR, cholesterol, blood pressure, antihypertensive or statin therapy, socioeconomic status, physical activity, inflammatory markers (serum interleukin [IL]-6, high-sensitivity C-reactive protein [hs-CRP]), and coronary artery calcification as a measure of subclinical atherosclerosis did not eliminate the association between %PredLVM and incident AVC (OR: 1.23/SD increase in %PredLVM [95% CI: 1.03 to 1.46]; p = 0.02) or incident MAC (OR: 1.19 [95% CI: 1.01 to 1.40]; p = 0.04).
We subsequently evaluated parameters of LV geometry. After adjusting for age, sex, BMI, ethnicity, study site, socioeconomic status and physical activity, %PredLVM and the LV mass/volume ratio were associated with both incident AVC and MAC (Table 3). Concentric LVH, defined by LV mass/volume ratio, was the strongest predictor of incident valve calcification (AVC: OR: 1.21/SD increase [95% CI: 1.03 to 1.43]; p = 0.02; MAC: OR: 1.27 [95% CI: 1.09 to 1.47]; p = 0.002]; whereas, there was no relationship between percentage of predicted LV volume and either AVC or MAC.
The association of increasing LV mass with the prevalence and severity of AVC is aligned with the paradigm that LVH develops in response to aortic valve disease. However, we also found associations between LV mass and the prevalence and severity of MAC, which does not increase LV afterload and has not been recognized as an independent cause of LVH. Furthermore, we demonstrate that increased LV mass, specifically concentric and noneccentric LVH, independently predicts the development of AVC and MAC over time. These observations suggest that concentric LVH might identify subjects at risk for valve calcification and possibly that common pathophysiological mechanisms might account for both the development of valvular calcification and LVH.
An obvious potential explanation for this association is hypertension, which has been consistently implicated as a factor in the development and progression of calcific valve disease (1,2,20) and is the most common cause of LVH in the general population (21). Hypertension, broadly defined on the basis of patient reporting, increased systolic or diastolic blood pressure on the initial study visit, or the use of any antihypertensive therapy was prevalent among 2,281 subjects in the MESA cohort (45%). Multivariable analyses adjusting for hypertension suggest that the observed relationships between LVH and valve calcification are independent of this factor; however, residual confounding by hypertension cannot be excluded. More sensitive measures of hypertension, such as ambulatory blood pressure monitoring, would be necessary to more definitively exclude hypertension as a mediator of the observed relationships.
Inflammatory and neurohormonal mechanisms have also been implicated in the development of both cardiovascular calcification and LVH. Angiotensin II and several inflammatory cytokines are involved in both myocardial remodeling and valve calcification, and any or all of these could mediate the relationship between %PredLVM and incident valve calcification (1,2,22–25). Exploring potential links, we adjusted for cardiovascular risk factors, the inflammatory markers hs-CRP and IL-6, and subclinical atherosclerosis by sequential multivariable analyses and found no attenuation of the association between LVH and valve calcification, suggesting that none of these factors entirely explains the relationship. However, we did not possess measures of other neurohormonal and inflammatory pathways. Alternatively, valve remodeling, inflammation, and calcification have been linked to changes in the hemodynamic environment. Conceivably, LVH-associated alterations in shear stress and cyclic pressures and stretch might induce or propagate valve pathology (26,27). Regardless of the etiology, our analyses suggest that concentric LVH identifies those subjects at risk for developing valve calcification.
First is the limited power to fully evaluate factors that are associated with valve calcification and LVH, given the relative good health of the MESA cohort, of which only 5% developed incident valve calcification. This restriction also hinders the ability to fully characterize the apparent nonlinear relationship between %PredLVM and valve calcification. Second is the limited data on duration and severity of hypertension, which precludes a more robust adjustment for hypertension. Third is the relatively modest relationship between %PredLVM and incident valve calcification. Despite these limitations, the association between LVH, specifically concentric remodeling, and valve calcification has potentially important clinical implications and warrants further study.
We found in the diverse MESA cohort that LVH is associated with the prevalence and severity of calcification of the aortic and mitral valves. Moreover, increased LV mass, specifically concentric LVH, at baseline was associated with the risk of incident AVC and MAC. These associations were independent of hypertension or other atherosclerotic risk factors, hs-CRP, IL-6, and subclinical atherosclerotic disease, suggesting that LVH might identify subjects at risk of developing valve calcification. Further study is needed to determine the pathophysiological links involved and to evaluate their reversibility and impact on patient outcomes.
The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.
This work was supported by the National Heart, Lung, and Blood Institute at the National Institutes of Health (Grants R01 HL071739 [to Dr. Budoff] and T32 HL007824 [to Dr. Elmariah] and by contracts N01-HC-95159 through N01-HC-95165 and N01-HC-95169) and the GlaxoSmithKline Research and Education Foundation for Cardiovascular Disease International Competitive Grants Award Program for Young Investigators (to Dr. Elmariah). Dr. Budoff has a modest consulting agreement with the General Electric Company. Dr. O'Brien receives speaker honoraria from AstraZeneca and Merck. Dr. Fuster chairs the HRP study, which is funded by BG Medicine. Dr. Halperin receives consulting fees from Astellas Pharma, Bayer AG Healthcare, the Bristol-Myers Squibb/Sanofi Partnership, Boehringer-Ingelheim, Daiichi Sankyo, Johnson & Johnson, and Sanofi-Aventis; honoraria from Genzyme and Portola Pharmaceuticals; and is co-chairman of the IMPACT trial, which is sponsored by Biotronik. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. H. William Strauss, MD, served as Guest Editor for this paper.
- Abbreviations and Acronyms
- percentage of predicted left ventricular mass
- percentage of predicted left ventricular end diastolic volume
- Agatston units
- aortic valve calcification
- body mass index
- confidence interval
- high-sensitivity C-reactive protein
- left ventricular
- left ventricular hypertrophy
- mitral annulus calcification
- odds ratio
- Received August 11, 2011.
- Revision received December 8, 2011.
- Accepted December 15, 2011.
- American College of Cardiology Foundation
- Goldbarg S.H.,
- Elmariah S.,
- Miller M.A.,
- Fuster V.
- Atar S.,
- Jeon D.S.,
- Luo H.,
- Siegel R.J.
- Barasch E.,
- Gottdiener J.S.,
- Marino Larsen E.K.,
- Chaves P.H.,
- Newman A.B.
- Bild D.E.,
- Bluemke D.A.,
- Burke G.L.,
- et al.
- Agatston A.S.,
- Janowitz W.R.,
- Hildner F.J.,
- Zusmer N.R.,
- Viamonte M. Jr..,
- Detrano R.
- Bluemke D.A.,
- Kronmal R.A.,
- Lima J.A.,
- et al.
- Bertoni A.G.,
- Whitt-Glover M.C.,
- Chung H.,
- et al.
- Elmariah S.,
- Delaney J.A.,
- O'Brien K.D.,
- et al.
- O'Brien K.D.,
- Shavelle D.M.,
- Caulfield M.T.,
- et al.
- Jiang Y.,
- Kohara K.,
- Hiwada K.