Hypertension has been identified as a risk factor for aortic valve calcium (AVC) but the magnitude of the risk relation with hypertension severity or whether age affects the strength of this risk association has not been studied. The relation of hypertension severity, as defined by Joint National Committee 7 (JNC-7) hypertension stages or blood pressure (BP), to computed tomographically assessed AVC prevalence and severity was examined in 4,274 participants in the Multi-Ethnic Study of Atherosclerosis (MESA) without treated hypertension. Analyses were stratified by age <65 or ≥65 years, were adjusted for common cardiovascular risk factors, and excluded those on antihypertensive medications. In age-stratified adjusted analyses, stage I/II hypertension was associated with prevalent AVC in those <65 but not in those ≥65 years of age (odds ratio [OR] 2.31, 95% confidence interval [CI] 1.35 to 3.94, vs 1.33, 0.96 to 1.85, p for interaction = 0.041). Similarly, systolic BP and pulse pressure were more strongly associated with prevalent AVC in those <65 than in those ≥65 years of age (OR 1.21, 95% CI 1.08 to 1.35, vs 1.07, 1.01 to 1.14, per 10-mm Hg increase in systolic BP, p for interaction = 0.006; and OR 1.41, 95% CI 1.21 to 1.64, vs 1.14, 1.05 to 1.23, per 10-mm Hg increase in pulse pressure). No associations were found between hypertension stage or BP and AVC severity. In conclusion, stage I/II hypertension and higher systolic BP and pulse pressure were associated with prevalent AVC. These risk associations were strongest in participants <65 years of age.
Previous studies evaluating the association of hypertension with aortic valve disease have used various definitions of hypertension and mostly have dichotomized hypertension as present or absent in adjusted risk models, rather than examining blood pressure (BP) as a continuous risk variable. Also, none have examined the association of aortic valve calcium (AVC) with various stages of hypertension as defined by Joint National Committee 7 (JNC-7) guidelines. Furthermore, most studies evaluating this relation have used echocardiography to assess AVC. Recently, computed tomography has been established as a highly reproducible method for quantifying AVC. Severity of AVC has been shown to be an independent predictor of worse clinical outcomes. We hypothesized that BP components and hypertension stages are independent predictors of the presence and severity of AVC, as assessed using baseline computed tomographic (CT) scans from patients enrolled in the Multi-Ethnic Study of Atherosclerosis (MESA). In addition, we investigated if age modified these associations, similar to previously observed attenuation of risk for AVC with dyslipidemia in advancing age.
Methods
The MESA is a prospective cohort study designed to investigate the prevalence and progression of subclinical cardiovascular disease in a population-based sample. Details of the study design have been published elsewhere. There were 6,814 participants in MESA 45 to 84 years of age from 4 ethnic groups, i.e., white, African-American, Hispanic, and Chinese. Recruitment took place from July 2000 to August 2002 at 6 United States field centers (Baltimore City and Baltimore County, Maryland; Chicago, Illinois; Forsyth County, North Carolina; Los Angeles County, California; New York, New York; and St. Paul, Minnesota). All participants were free of known cardiovascular disease at time of enrollment. Institutional review boards at all participating institutions approved the study, and all participants provided informed consent.
Baseline information including demographic variables, socioeconomic status, medical history, and medications were collected using questionnaires. During the initial examination, height and weight and waist and hip circumferences were measured. BP at rest was measured 3 times in the seated position using a Dinamap model Pro (GE Critikon, Wauwatosa, Wisconsin) 100 automated oscillometric sphygmomanometer. The last 2 measurements were averaged and used for these analyses.
Chest computed tomography was performed in all participants 2 times during the initial examination. Images were generated with a cardiac-gated electron beam CT scanner (Imatron C-150; GE Medical Systems, Milwaukee, Wisconsin) or electrocardiogram-triggered 4-detector row CT scanners. Details of scanning method and quality control, reproducibility, and effect of scanner type in the MESA study have been reported previously. AVC was defined as any calcified lesion within the aortic valve leaflets, excluding lesions involving the aortic and mitral annulus, sinuses, and wall of the ascending aorta. AVC quantification of lesions was determined using the method of Agatston et al. A total AVC score was calculated for each participant from the summation of all individual lesions. Absence of AVC was assigned a score of 0. All studies were evaluated retrospectively at the MESA CT reading center (Harbor-UCLA Research and Education Institute, Los Angeles, California) by a single blinded reader (J.T.).
Comparisons between BP categories based on JNC-7 criteria to demographic measurements and cardiovascular risk factors are expressed using means and proportions. AVC scores first were categorized as present (AVC >0) or absent (AVC 0). We used a 2-stage regression approach for analysis. Because prevalence of calcification is <10% we use odd ratios (ORs) from a logistic regression model to estimate relative risks. Using this method we examined prevalence of AVC >0 (i.e., proportion with scores >0) in unadjusted and adjusted models with BP stage categories, based on JNC-7 criteria, and continuous measurements of systolic BP, diastolic BP, and pulse pressure. In those with detectable calcium (i.e., in participants with strictly AVC >0) we used the AVC score as a continuous variable in a multivariable linear regression to assess associations of AVC severity with BP. Given the skewed nature of the AVC distribution, natural log transformation of AVC was modeled. Models for logistic and linear regressions were adjusted for age, gender, race, body mass index, smoking, diabetes, low-density lipoprotein concentration, high-density lipoprotein concentration, and use of lipid-lowering medications.
Age interactions were tested in adjusted regression analyses and considered statistically significant at a p value <0.05. Age was stratified as >65 or <65 years to compare to previous elderly cohort analysis. Because there was a small number of younger participants with advanced hypertension, stages I and II were combined for these analyses. If interactions were significant, we presented the stratified results.
We excluded participants with treated hypertension (n = 2,533), as done in previous BP studies in MESA, because the effect of BP medication on AVC and the amount of misclassification because of treatment of BP are unknown. A sensitivity analysis including treated hypertensive patients was conducted but did not change inferences. All statistical analyses were performed with SPSS 16.0.1 for Windows (SPSS, Inc., Chicago, Illinois) and S-Plus 8.0 (Insightful, Inc., Seattle, Washington).
Results
There were 4,275 participants without treated hypertension in this analysis, and 2,783 were <65 years old. Differences were present in AVC risk factors between those with and without hypertension ( Table 1 ). Hypertensive participants were more often older, diabetic, had a higher body mass index, and more likely to be African-American. Overall, 407 participants had AVC, for a prevalence of 9.5% in this cohort. This AVC prevalence is lower than in previous reports from the MESA cohort due to exclusion of participants with treated hypertension who had a prevalence of AVC of 20%.
| Variable | Normal (n = 2,206) | Borderline Hypertension (n = 1,304) | Stage I (n = 566) | Stage II (n = 199) |
|---|---|---|---|---|
| Aortic valve calcium prevalence | 139 (6%) | 144 (11%) | 93 (17%) | 31 (16%) |
| Aortic valve calcium score † | 45 (17–144) | 54 (18–115) | 73 (15–209) | 69 (17–246) |
| Age (years) | 57 ± 9 | 62 ± 10 | 65 ± 10 | 68 ± 9 |
| Women | 1,215 (55%) | 600 (46%) | 279 (49%) | 118 (59%) |
| White | 960 (44%) | 515 (40%) | 215 (38%) | 63 (32%) |
| Chinese | 325 (15%) | 153 (12%) | 71 (13%) | 23 (12%) |
| Black | 402 (18%) | 320 (25%) | 148 (26%) | 70 (35%) |
| Hispanic | 519 (24%) | 316 (24%) | 131 (23%) | 43 (22%) |
| Ever smoked | 1,079 (49%) | 662 (51%) | 297 (53%) | 98 (50%) |
| Body mass index (kg/m 2 ) | 26.8 ± 5.0 | 28.4 ± 5.1 | 28.0 ± 5.3 | 29.2 ± 6.0 |
| Waist circumference (cm) | 93 ± 14 | 99 ± 13 | 98 ± 14 | 101 ± 15 |
| Diabetes mellitus | 140 (6%) | 124 (10%) | 69 (12%) | 26 (13%) |
| Glucose (mg/dl) | 98 ± 26 | 104 ± 33 | 104 ± 28 | 106 ± 27 |
| Low-density lipoprotein (mg/dl) | 119 ± 32 | 121 ± 31 | 121 ± 30 | 123 ± 32 |
| High-density lipoprotein (mg/dl) | 52 ± 15 | 51 ± 15 | 52 ± 15 | 52 ± 15 |
| Triglycerides (mg/dl) † | 103 (72–148) | 115 (77–165) | 119 (82–168) | 116 (85–165) |
| Lipid-lowering medications | 206 (9%) | 141 (11%) | 70 (12%) | 18 (9%) |
⁎ Joint National Committee 7 categories: normal, systolic blood pressure <120 mm Hg and diastolic blood pressure <80 mm Hg; borderline hypertension, systolic blood pressure 120 to 139 mm Hg or diastolic blood pressure 80 to 89 mm Hg; stage I, systolic blood pressure 140 to 159 mm Hg or diastolic blood pressure 90 to 99 mm Hg; stage II, systolic blood pressure ≥160 mm Hg or diastolic blood pressure ≥100 mm Hg.
Adjusted odds ratio (OR) for stage I or II hypertension was 1.27 (95% confidence interval [CI] 0.96 to 1.69). In multivariable analyses, there was a significant interaction between age and hypertension stage (p for interaction = 0.041). Stage I or II hypertension was more strongly associated with prevalent AVC in participants <65 years of age than in those ≥65 years of age ( Figure 1 ). JNC-7–defined borderline hypertension was not significantly associated with an increased OR for prevalent AVC in either age stratum.
We found similar significant age interactions with continuous measurements of BP (p for interaction = 0.006). Although adjusted ORs for prevalent AVC in systolic BP (per 10-mm Hg increase) were increased significantly for participants <65 and those ≥65 years of age ( Figure 2 ), the association was significantly greater in those <65 years of age. There was no association of prevalent AVC with diastolic BP ( Figure 2 ). Of BP variables, pulse pressure had the strongest association with prevalent AVC ( Figure 2 ).
There were no statistically significant associations or age interactions of JNC-7–defined hypertension stages ( Table 2 ) or BP ( Table 3 ) with AVC severity.
| Participants With AVC >0 | Unadjusted Exp(β) † (95% CI) | Adjusted ⁎ Exp(β) † (95% CI) | |
|---|---|---|---|
| Normal | 133 | 1.00 (reference) | 1.00 (reference) |
| Borderline hypertension | 144 | 1.07 (0.75–1.52) | 0.91 (0.65–1.29) |
| Stages I and II | 125 | 1.33 (0.92–1.92) | 1.18 (0.82–1.70) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
