Blood Pressure and Vascular Calcification

Epidemiological studies suggest that elevated blood pressure (BP) is an independent and strong predictor of cardiovascular disease (CVD).¹ Controversy exists as to which BP measure is the best predictor of CVD events. Recent data suggest that aortic stiffness is an independent predictor of future heart disease in older individuals.² Increases in aortic stiffness are believed to be closely linked to increases in pulse pressure, thereby placing a higher afterload on the left ventricle.³ Thus, pulse pressure (PP) may be relevant in the pathophysiology of coronary heart disease.⁴

Calcium is deposited early in the formation of the atherosclerotic plaque and can be used as a marker of the atherosclerotic process.⁵ The use of electron beam computed tomography affords the opportunity to noninvasively construct cross-sectional images of arteries to detect the presence and extent of calcium attributed to atherosclerosis in different vascular beds.⁶ The coronary calcium score is significantly predictive of future cardiac events.⁷

There are many risk factors associated with atherosclerotic calcification, including hypertension (HTN).⁸ Previous reports have not typically explored the association between different measures of BP and calcification in multiple vascular beds. Accordingly, the aim of the study was to determine which measures of BP (systolic BP [SBP], diastolic BP [DBP], mean arterial pressure [MAP], PP, HTN, and isolated systolic HTN [ISH]) were associated with arterial calcification in 6 different vascular beds (carotids, coronaries, subclavians, thoracic aorta, abdominal aorta, and iliacs).

Methods

Subjects

From October 1999 to July 2003, 9510 ambulatory patients presented for preventive medicine services at a private, university-affiliated disease prevention center in La Jolla, California. Most patients were asymptomatic and self-referred or referred by their primary physician as a supplement to their routine medical care. All of the patients completed a detailed health history questionnaire before undergoing the whole body scanning procedure. Smoking status was defined as current, former, or never. Family history of coronary heart disease (CHD) was defined as a CHD event before age of 55 years in a first-degree relative. Patients who had a history of CHD-related surgery (stent placement or coronary artery bypass graft) were excluded from the study.

The study protocol complies with the Declaration of Helsinki and was approved by the committee for protection of human subjects at the University of California San Diego, which granted a waiver for informed consent because we analyzed existing data from a clinical population.

Imaging

All 9510 patients had electron beam computed tomography scans of the heart with an Imatron C-150 scanner (GE). Of these, 6456 had “whole body scans,” which entailed scanning from the base of the skull to the pubic symphysis. From this scan, we obtained measurements of calcified atherosclerosis in the following vascular beds: carotids, subclavians, thoracic aorta, abdominal aorta, and iliac arteries, as described previously.⁹ The coronary and carotid images were analyzed at the time of the electron beam computed tomography scan. For the other vascular beds, images were retrospectively analyzed for calcified atherosclerosis. Calcium scoring were performed using the method described by Agatston et al.¹⁰ The coronary calcium score consisted of calcified lesions in the left main, left anterior descending, left circumflex, and right coronary arteries. Data from the left and right sides were combined to give the extent of calcium in the carotid, subclavian, and iliac beds. The thoracic aorta was defined as the segment above the diaphragm, whereas the abdominal aorta was the segment below the diaphragm to the aortic bifurcation. The right and left subclavian arteries were evaluated from their takeoff points from the brachiocephalic artery and aorta, respectively, to the apex of the vessels as they began to arch over the top of the lung.

Blood Pressure

After the patient had rested for 5 minutes in the seated position, trained technicians used a standardized protocol to obtain SBP and DBP in the right upper extremity using a stethoscope and a sphygmomanometer. BP was assessed once at the clinic visit. PP was calculated as the difference between the SBP and DBP, whereas MAP was defined as DBP +1/3 (SBP−DBP). ISH was defined as an SBP of ≥140 mm Hg and a DBP <90 mm Hg and no reported use of HTN medication. HTN was defined as SBP ≥140 and DBP ≥90 and/or use of HTN medication.

Laboratory

Casual serum lipid and glucose measurements were obtained by finger stick using the Cholestec LDX system. Weight was assessed with the patient lightly clothed and without shoes. Body mass index (BMI) was calculated as the weight (in kilograms) divided by the height (in meters squared). Body fat measurement was conducted using bioimpedance on the OMRON HBF-300.

Diabetes mellitus was defined by current use of physician-prescribed antiglycemic medications or a casual glucose level >200 mg/dL. Individuals with a total:high-density lipoprotein cholesterol ratio >5 or those who reported the use of a medication to treat high cholesterol were classified as dyslipidemic.

Statistical Analysis

The outcome variables for this study were the presence and extent of calcium attributed to atherosclerosis in the carotids, subclavians, coronaries, thoracic aorta, abdominal aorta, and iliac arteries. Six measures of BP were the primary predictor variables: SBP, DBP, PP, MAP, ISH, and HTN. Covariates included age, sex, BMI, smoking, dyslipidemia, diabetes mellitus, family history of CHD, and HTN medication. HTN medication was included in the definition of ISH and HTN.

The differences in the distributions of the baseline characteristics were determined by t tests or χ² tests, as appropriate. For the primary analysis, calcium scores were dichotomized as either the presence or absence of calcified plaques, and logistic regression analysis was used to determine odds ratios. In analyses limited to those with calcium in the given vascular bed, multivariable linear regression was used to determine the associations between measures of BP and the extent of vascular calcium in distinct vascular beds. To make β-coefficients more interpretable and units comparable, all of the BP parameters used for statistical analyses were defined by dividing the means of the PP, SBP, DBP, and MAP by their SD.

We established a priori that the relationship between age and BP measures, as well as sex and BP measures, may potentially influence the outcome. Therefore, we stratified the cohort by age (≥60 and <60 years) and separately for sex and repeated our analyses as described above. For each outcome variable (calcium scores in each vascular bed), we did not impute data but rather used the lowest common denominator of subjects for each analysis after accounting for missing values. For example, subclavian calcium scores were analyzed in 1477 subjects, whereas coronary calcium was analyzed in 9510 subjects. All of the P values were 2 tailed, with a P value of <0.05 considered to indicate statistical significance. All of the statistical analyses were conducted using SPSS (version 16.0; SPSS, Inc).

Results

The total number of subjects examined for presence of calcium in the carotids, coronaries, subclavians, thoracic aorta, abdominal aorta, and iliacs were 4487, 9510, 1477, 4606, 4606, and 4604, respectively. The number of subjects varied because the images of the thoracic and abdominal aortas, as well as the subclavians and iliacs, were analyzed retrospectively. This, combined with differences in resources available during the time frames when each of the vascular beds was analyzed, led to the different number of analyzed scans for each bed.

The characteristics of all of the study subjects (n=9510) are provided in Table 1. Men composed 57.5% of the study sample. More than 20% of the sample had a family history of CHD, 37.5% had smoked or were currently smoking, 28.3% had HTN, 8.2% had ISH, and 2.6% had diabetes mellitus. The prevalence of any calcium was highest in the coronaries (57.2%), whereas the lowest prevalence was found in the carotid and subclavian arteries (31.9% and 31.7%, respectively).

After stratifying the cohort by 60 years of age, the ≥60 group had a higher SBP, PP, and MAP compared with the <60 group (P<0.01; Table 1). In addition, the ≥60 age group had a greater prevalence of ISH (12.1% compared with 5.8%; P<0.01). The prevalence of calcification was highest in the abdominal aorta followed by the iliacs and coronaries (84.2%, 73.9%, and 72.9%, respectively) for those ≥60 years of age. Conversely, the <60 age group had the greatest prevalence of calcification in the coronaries followed by the abdominal aorta (45.8% and 34.6%, respectively).

Table 2 shows the age-adjusted differences in patient characteristics between those patients with calcium compared with patients without calcium in different vascular beds. In general, there were significant differences in patient characteristics between those with and without calcium. The differences between the 2 groups were often seen in age, SBP, PP, smoking, and HTN. In the coronaries, there were significant differences between the groups for all of the characteristics listed.

Table 3 shows the odds of having calcium scores >0. In separate multivariable logistic models containing traditional CVD risk factors (age, sex, BMI, smoking, dyslipidemia, diabetes mellitus, family history of CVD, and HTN medication), PP and SBP were significantly associated with calcification in all of the vascular beds except the iliacs and subclavians, respectively, with PP having the stronger magnitudes of the associations for most of the vascular beds. Similarly, and compared with HTN, a diagnosis of ISH was associated with a greater odds of calcification in a majority of the vascular beds.

These analyses were repeated and stratified by age (Table 3). In the ≥60 group, there were significant associations between both SBP and PP and vascular bed calcification in all of the different sites. In general, the odds for the presence of calcium for both PP and SBP were greater in the ≥60 group compared with the <60 group. Lastly, compared with those without ISH, those with ISH had a greater magnitude of association with calcified atherosclerosis among all of the vascular beds in the ≥60 group compared with the <60 group. In this regard, there were significant interactions between age and PP, SBP, and ISH for all of the vascular beds (P≤0.01 for all).

Table 4 illustrates sex differences for the odds of calcium >0. These data, suggest that men have a greater odds of calcification in all of the vascular beds compared with women. In addition, higher PP, as well as a diagnosis of ISH and HTN, was associated with a higher odds of calcification in all of the vascular beds. Notably, in women, ISH appears to be more strongly associated with calcification, whereas, in men, HTN is associated with a higher odds of calcification to a greater magnitude. There were significant interactions between sex and BP variables (SBP, PP, ISH, and HTN).

Multivariable linear regression models were used to determine the associations with the extent of calcium in each vascular bed (Table 5) among subjects with any calcification. A 1-SD increase in SBP and PP corresponded with the greatest increases in the extent of calcium in the thoracic aorta. In addition, a diagnosis of ISH was associated with the largest magnitudes of the associations for increasing calcium in the thoracic aorta and coronaries, whereas HTN had the largest magnitudes of associations in the subclavians and carotids (Table 5). In a majority of cases, these increases in association tended to be greater in the ≥60 group. In these analyses, there were significant interactions between age and the following BP variables: SBP, PP, and ISH (P≤0.01 for all).

We also determined the associations by sex among subjects with any calcification (Table 6). The results were similar to the previous findings in that SBP and PP corresponded with the greatest increases in the extent of calcium in the thoracic aorta for continuous variables in both men and women, and there were significant interactions between sex and the same BP variables.

Discussion

Before our study, there was only limited information on the presence of atherosclerotic calcification in noncoronary vascular beds.¹¹ A previous study demonstrated that the detection of extracoronary atherosclerosis potentially helps to identify individuals more likely to develop new calcified coronary disease.¹² Furthermore, controversy exists as to which BP components are the best predictors of CVD events and which components of BP are most strongly associated with calcium in noncoronary vascular beds. Our study is unique in that we analyzed the associations between multiple measures of BP and calcification in distinct vascular beds. Overall, for continuous variables, we found that SBP and PP had the greatest odds of calcification across all vascular beds. Furthermore, the BP parameters SBP and PP corresponded with the greatest magnitude of associations in the thoracic aorta. ISH had a greater magnitude of association in the thoracic aorta and coronaries when compared with a diagnosis of HTN.

Previous studies have determined that BP components have age-dependent roles in the prediction of coronary artery calcification¹³ and that SBP, DBP, MAP, and PP¹⁴ are age-dependent predictors of CVD risk.¹⁵ In this regard, with increasing age, arterial compliance decreases and arterial stiffness increases,¹⁵ leading to increases in SBP, decreases in DBP, and a widening of PP.¹⁶ In our study and among those ≥60 years of age, SBP and PP were elevated, and more subjects had ISH. Conversely, in those <60 years of age, HTN had stronger associations with calcification compared with ISH, and both HTN and ISH were more variable in women. Similarly, in the Rochester Family Heart Study, SBP and PP were independent predictors of the presence and quantity of coronary artery calcification in the ≥50-years-of-age group.¹³ Taken together, these results suggest that the relevance of PP and ISH (and perhaps SBP) increases with age, particularly in the sixth and seventh decades of life.

Many epidemiological studies have focused on the presence of significant associations between PP and coronary atherosclerosis¹⁷ and determined that PP was more highly correlated with atherosclerosis when compared with SBP or DBP levels. Furthermore, previous studies indicated that SBP and PP may be better predictors of CVD risk than other measures of BP.¹⁸ Our data indicate that PP is significantly associated with vascular calcification in other noncoronary vascular beds, except the iliacs and subclavians, and the magnitudes of associations are generally stronger than for SBP and DBP.

Atherosclerotic calcification may lead to arterial stiffness, a significant risk factor for CVD.¹⁹ Notably, ISH reflects an elevated PP, which is considered a crude estimate of arterial stiffness, which is a powerful predictor of cardiovascular events.²⁰ Our results for these 2 measurements showed significant associations with calcium in the thoracic aorta, and the magnitudes of the associations differed significantly by age group. Notably, in the elderly, ISH is associated with increased aortic stiffness, early return of reflected waves, and, as a result, smaller PP amplification.²¹ Our results support the relevance of PP and ISH in the pathophysiology of CVD in those ≥60 years of age.

Individuals with ISH have an increased risk for stroke, CHD, and congestive heart failure.²² The prevalence of ISH increases among older subjects, as does the magnitude of the association of ISH with calcification and aortic stiffness.²³ Importantly, studies suggest that aortic calcification may play an important role in the development of ISH in humans, as is the case in animal models of vascular calcification,²⁴ in which drug-induced vascular calcification leads to the development of ISH.²⁵ In this way, vascular calcification may be relevant in the development of ISH rather than ISH leading to calcification. Clearly, prospective studies on the association between aortic calcification and ISH in humans are needed to clarify this relationship and to determine whether HTN promotes calcifications or whether calcifications promote stiffness and, therefore, increases in SBP and PP.

The strengths of our study include the large sample size, the ascertainment of calcification measures in multiple vascular beds, and the use of multiple measures of BP for analytic purposes. Our study has some limitations. The results of this study may not be generalizable to all community-based populations, because the subjects for this study were self-selected. However, the 10-year Framingham risk score from our study cohort was comparable to the Framingham cohort,²⁶ which was similar in ethnic distribution to ours. For instance, after calculating the 10-year Framingham risk score for our study cohort, we determined that men had a 10% risk of developing CHD and women had a 2% risk of developing CHD, which was similar to the original cohort where men had a risk of 8.0% followed by women who had a risk of 2.8%. In addition, small differences in our data resulted in significance, presumably because of our large sample size; thus, significance may not ensure clinical relevance. Another potential limitation is that computed tomography cannot determine whether calcium is in the intima or media of the arterial wall. Calcium in the intimal layer of the artery is associated with atherosclerotic plaque and is a predictor of cardiac events.²⁷ In contrast, Mönckeberg sclerosis, or calcification of the tunica media layer, is associated with metabolic, electrolyte, and pH abnormalities. Mönckeberg calcification is also frequently associated with renal disease or diabetes mellitus.²⁸ However, we had few subjects with diabetes mellitus in our study, and, thus, we believe that the possibility of misclassification is small. Also, although kidney function influences BP, history of chronic kidney disease was not collected. Importantly, previous studies have not found a significant association between kidney function and vascular calcification.²⁹

In conclusion, the current study demonstrates that different measures of BP are associated with significant calcification in multiple vascular beds and that these associations vary with age. Our study also highlights the differences between the more distal vascular beds (iliacs and subclavians) versus the more central vascular beds (thoracic and abdominal aorta). That is, PP and SBP, as well as a diagnosis of ISH, had the greatest magnitudes of association in the thoracic aorta. From a BP perspective, the vascular beds studied appear to be physiologically distinct, which may potentially explain these differences. We also assessed the extent of calcification at multiple sites to test the relation of the number of diseased sites with BP measures. We determined that there were no significant findings after adjusting for CVD risk factors, which may imply bed specificity in the BP-calcification relationship. The results presented underscore the importance of assessing PP and ISH in the prevention of CVD.

Perspectives

Potentially, our findings may be used to optimize clinical protocols for the diagnosis of atherosclerosis and the prevention of ischemic heart disease and stroke. Also, measuring multiple vascular beds may be a reasonable risk-reduction strategy to reduce CVD. Hopefully, data from future studies will provide additional insight that suggests that screening multiple vascular beds may have clinical relevance and that screening multiple BP components will be useful in improving CVD risk assessment.

Footnotes