Cross-Sectional Relations of Arterial Stiffness, Pressure Pulsatility, Wave Reflection and Arterial Calcification

Abstract

Objective

Arterial hemodynamics and vascular calcification are associated with increased risk for CVD, but their inter-relations remain unclear. We sought to examine the associations of arterial stiffness, pressure pulsatility, and wave reflection with arterial calcification in individuals free of prevalent cardiovascular disease (CVD).

Approach and Results

Framingham Heart Study Third Generation and Offspring Cohort participants free of CVD underwent applanation tonometry to measure arterial stiffness, pressure pulsatility, and wave reflection, including carotid-femoral pulse wave velocity (CFPWV), central pulse pressure (CPP), forward wave amplitude, and augmentation index (AI). Participants in each cohort (n=1905, 45±6 years and n=1015, 65±9 years, respectively) underwent multi-detector computed tomography to assess presence and quantity of thoracic (TAC) and abdominal (AAC) aortic calcification and coronary artery calcification (CAC). In multivariable-adjusted models, both higher CFPWV and CPP were associated with greater TAC and AAC, whereas higher AI was associated with AAC. Among the tonometry measures, CFPWV was the strongest correlate of all calcification measures in multivariable-adjusted models (odds ratio [OR] per SD for TAC 2.69 (95%CI 2.17-3.35), AAC 1.47 (95%CI 1.26-1.73), and CAC 1.48 (95%CI 1.28-1.72), all p<0.001, respectively). We observed stronger relations of CFPWV, CPP, and forward wave amplitude with nearly all continuous calcification measures in the younger Third Generation Cohort as compared with the Offspring Cohort.

Conclusions

In community-dwelling individuals without prevalent CVD, abnormal central arterial hemodynamics were positively associated with vascular calcification, and were observed at younger ages than previously recognized. The mechanisms of these associations may be bidirectional and deserve further study.

INTRODUCTION

Higher arterial stiffness and pressure pulsatility, hallmarks of aging, are associated with increased risk for cardiovascular disease (CVD) morbidity, including incident hypertension.¹ Carotid-femoral pulse wave velocity (CFPWV), the current reference standard measure of global aortic stiffness, is associated with CVD events including coronary heart disease, stroke, and CVD mortality.^2-5 Arterial calcification, in turn, is also associated with CVD events and mortality, whether assessed in the coronary arteries (CAC),^{6, 7} thoracic aorta (TAC),^{8, 9} or abdominal aorta (AAC)¹⁰. Early studies noted greater arterial stiffness and arterial calcification in older people, suggesting a possible association between these two pathological processes.^{11, 12}

Evidence from experimental studies¹³ and clinical reports ^14-20 has suggested that the processes of arterial stiffening and calcification likely do not develop in parallel, but rather share mechanisms that are incompletely understood. Arterial stiffening is thought to involve changes in the extracellular matrix, including degradation of elastin and deposition of collagen, and may be exacerbated by arterial calcification.²¹ However, recent evidence suggests that vascular remodeling in the presence of increased stiffness may contribute to medial and intimal calcification.²²

We hypothesized that tonometric measures of aortic stiffness (CFPWV), pressure pulsatility (central pulse pressure, CPP), and wave reflection are associated positively with segmental calcification of the thoracic and abdominal aorta as well as the coronary arteries. Using a cross-sectional design, we sought to test our hypotheses in a community-based sample without prior overt CVD.

MATERIALS AND METHODS

Materials and Methods are available in the online-only Data Supplement.

RESULTS

Characteristics of the FHS Offspring and Third Generation Cohorts are presented in Table 1. The older Offspring Cohort manifested the expected higher prevalences of age-related characteristics, including hypertension, use of anti-hypertensive and lipid-lowering medications, and diabetes. The Offspring also had greater arterial stiffness and pulsatility measures and a higher prevalence and extent of arterial calcification. The characteristics of FHS participants who were excluded for incomplete MDCT measures are presented in Supplemental Table I. Compared to study participants, Third Generation participants without MDCT measures were slightly shorter and had a greater prevalence of smoking. Offspring Cohort participants with incomplete MDCT measures had greater CVD risk factors and tonometry measures than their study counterparts.

Table 1.

Characteristics of FHS Third Generation and Offspring Cohorts (n=2920)

Characteristic	Third Generation (n=1905)	Offspring (n=1015)	P-value
Age (yr)	44.9±6.2	65.2±8.7	<0.001
Male (%)	55	44	<0.001
BMI (kg/m2)	27.2±6.0	28.0±5.1	<0.001
Height (in)	67.6±3.7	65.8±3.8	<0.001
Systolic blood pressure (mm Hg)	119±17	127±16	<0.001
Diastolic blood pressure (mm Hg)	77±9	74±9	<0.001
Mean arterial pressure (mm Hg)	92±11	97.5±11	<0.001
Total Cholesterol (mg/dl)	194±34.5	190±35	0.005
HDL Cholesterol (mg/dl)	54.0±16.9	57.9±17.5	<0.001
Prevalent hypertension (%)	20	53	<0.001
Hypertension medications (%)	11	44	<0.001
Lipid lowering medications (%)	10	40	<0.001
Diabetes (%)	3	12	<0.001
Current smoking (%)	14	7	<0.001
Tonometry measures
Heart rate (bpm)	61±9	61±10	0.075
CFPWV (m/s)	7.4±1.3	10.1±3.4	<0.001
Central pulse pressure (mm Hg)	51.0±12.5	68.6±20.7	<0.001
Forward wave amplitude (mm Hg)	44.2±10.2	56.6±16.2	<0.001
Augmentation index (%)	9.7±12.9	15.6±11.9	<0.001
Augmented pressure (mm Hg)	6.5±6.6	11.7±9.6	<0.001
Reflection coefficient (%)	0.35±0.06	0.36±0.06	<0.001
Reflection wave amplitude (mm Hg)	15.2±3.9	20.2±6.0	<0.001
Prevalent calcification
TAC (%)	5	45	<0.001
AAC (%)	34	78	<0.001
CAC (%)	28	66	<0.001
Calcification, continuous
Loge (TAC+1), median (Q1, Q3)	0 (0, 0)	0 (0, 4.66)	<0.001
Loge (AAC+1), median (Q1, Q3)	0 (0, 3.21)	5.95 (1.86, 7.55)	<0.001
Loge (CAC+1), median (Q1, Q3)	0 (0, 0.68)	2.94 (0, 5.13)	<0.001

AAC: abdominal aortic calcium. CAC: coronary artery calcium. CFPWV: carotid-femoral pulse wave velocity. TAC: thoracic aortic calcium.

CPP and forward wave amplitude were highly correlated and other tonometry measures were modestly but significantly correlated (Supplemental Table IIa). In the Third Generation cohort, CFPWV was positively associated with AI. However, this relation reversed in the Offspring Cohort. Continuous calcification measures were moderately correlated in the Third Generation Cohort and the strength of association was greater in the Offspring (Supplemental Table IIb). We observed significant relations of arterial stiffness and pressure pulsatility with continuous measures of TAC, AAC, and CAC (Table 2). In both multivariable models adjusting for CVD risk factors without or with the addition of SBP, CFPWV was associated positively with all three measures of arterial calcification. The magnitude of associations of CFPWV and CPP (modeled as continuous variables) with arterial calcifications in the territories evaluated followed a similar regional pattern of TAC > AAC > CAC (p values, Table 2). Forward wave amplitude and AI were primarily associated with continuous TAC and AAC, respectively. We noted effect modification by cohort status for the majority of relations between tonometry measures with continuous calcification measures (marked by asterisks in Table 2), with greater effect sizes in the younger Third Generation Cohort (Supplemental Table III). The pattern of associations of CFPWV along the spectrum of TAC, AAC, and CAC were continuous (Figures 1A-C), and tests for non-linearity were not significant (p>0.05).

Table 2.

Continuous relations of primary tonometry variables with vascular calcification in the pooled FHS Third Generation and Offspring Cohorts without prevalent CVD (n=2920)

Tonometry Measure	Model	Loge (TAC+1) B±SE	P value	Loge (AAC+1) B±SE	P value	Loge (CAC+1) B±SE	P value
CFPWV	MV MV+ SBP	1.93±0.18 2.03±0.19	<1.0*10−14 <1.0*10−14	0.84±0.11 0.78±0.12	8.0*10−14* 5.3*10−11*	0.73±0.11 0.68±0.11	5.5*10−12* 1.6*10−9*
Central pulse pressure	MV MV+ SBP	0.68±0.13 0.71±0.16	1.8*10−7* 8.0*10−6*	0.43±0.09 0.35±0.10	1.4*10−7* 8.1*10−4*	0.31±0.08 0.20±0.10	2.7*10−4* 3.3*10−2*
Forward wave amplitude	MV MV+ SBP	0.63±0.12 0.61±0.14	4.2*10−7* 1.3*10−5*	0.24±0.09 0.12±0.09	5.2*10−3* 1.9*10−1*	0.26±0.08 0.17±0.09	9.0*10−4* 4.9*10−2*
Augmentation Index	MV MV+ SBP	0.20±0.15 0.13±0.15	1.7*10−1 3.9*10−1	0.73±0.10 0.68±0.10	<1.0*10−14* 6.4*10−13*	0.23±0.09 0.17±0.09	5.8*10−3 5.1*10−2

CFPWV: carotid-femoral pulse wave velocity. CAC: coronary artery calcium. TAC: thoracic aortic calcium. AAC: abdominal aortic calcium. MV: multivariable. Log_e() represents the natural logarithm.

Multivariable models adjusted for age, sex, cohort, body mass index, height, heart rate, history of hypertension or lipid-lowering medications, total cholesterol/HDL, diabetes, and current smoking.

B±SE represents change in log_e (calcification measure +1) per SD higher value for the tonometry variable.

^*p<0.05 for cohort-interaction

Figure 1.

Associations of CFPWV with prevalent TAC (A), AAC (B), and CAC (C) in the Framingham Heart Study Offspring and Third Generation Cohorts. Restrictive cubic logistic regression splines were adjusted for age, sex, cohort, body mass index, SBP, history of hypertensive treatment, total cholesterol/HDL, history of lipid-lowering treatment, diabetes, and smoking. Knots were placed at the 5^th, 50^th, and 95^th percentiles. Wald tests for non-linear associations were non-significant (all p>0.05).

The relations of tonometry measures with prevalent TAC, AAC, and CAC were consistent with results of continuous calcification measures above. In multivariable models that included SBP, CFPWV was associated with the presence of calcification in all three regions while CPP was significantly associated with prevalent TAC and AAC (Table 3, all p<0.001). In individuals without prevalent aortic calcification, a significant relation between CFPWV and CAC persisted (OR 1.38, 95% CI, 1.08-1.75, p=0.009) whereas no relation was observed between CPP and CAC (p=0.59). Consistent with linear regression analyses, forward wave amplitude was associated most strongly with prevalent TAC and AI was associated only with prevalent AAC. In sensitivity analyses, similar results were obtained when presence of calcification was defined based on age- and sex-specific 90^th percentiles in a healthy referent group (Supplemental Table IV).

Table 3.

Relations of tonometry variables with prevalent arterial calcification in the pooled FHS Third Generation and Offspring Cohorts without prevalent CVD (n=2920)

Tonometry Measure	Model	Prevalent TAC OR (95% CI)	P value	Prevalent AAC OR (95% CI)	P value	Prevalent CAC OR (95% CI)	P value
CFPWV	MV MV + SBP	2.52 (2.06-3.08) 2.69 (2.17-3.35)	<0.001 <0.001	1.51 (1.30-1.75) 1.47 (1.26-1.73)	<0.001 <0.001	1.53 (1.33-1.76) 1.48 (1.28-1.72)	<0.001 <0.001
CPP	MV MV + SBP	1.39 (1.21-1.60) 1.43 (1.21-1.70)	<0.001 <0.001	1.35 (1.19-1.54) 1.33 (1.15-1.53)	<0.001 <0.001	1.20 (1.07-1.34) 1.13 (1.00-1.29)	0.002 0.06
Forward wave amplitude	MV MV + SBP	1.36 (1.18-1.55) 1.36 (1.16-1.59)	<0.001 <0.001	1.22 (1.07-1.38) 1.17 (1.03-1.34)	0.002 0.018	1.17 (1.05-1.31) 1.12 (0.99-1.26)	0.004 0.07
AI	MV MV + SBP	1.12 (0.96-1.30) 1.09 (0.93-1.27)	0.16 0.28	1.42 (1.26-1.61) 1.40 (1.23-1.58)	<0.001 <0.001	1.15 (1.03-1.29) 1.11 (0.99-1.25)	0.012 0.07

AAC: abdominal aortic calcium. Al: augmentation index. CAC: coronary artery calcium. CFPWV: carotid-femoral pulse wave velocity. CPP: central pulse pressure. TAC: thoracic aortic calcium.

Multivariable models adjusted for age, sex, cohort, body mass index, height, heart rate, history of hypertension or lipid-lowering medications, total cholesterol/HDL, diabetes, and current smoking.

OR represent odds for outcome of prevalent calcification per SD increment in tonometry measure.

Tests for all cohort interaction were NS (p>0.05).

In stepwise multivariable analysis selecting between CFPWV, CPP, forward wave amplitude, and AI in the model, CFPWV consistently entered the model. Of all the tonometry measures, CFPWV had the strongest association with prevalent calcification in each of the vascular territories (Table 4). CFPWV and CPP were jointly associated with prevalent TAC and AAC, and CFPWV alone was associated with prevalent CAC in the stepwise model. Measures of stiffness (CFPWV), pressure pulsatility (CPP) and wave reflection (AI) each contributed separately to risk of prevalent AAC (Table 4).

Table 4.

Tonometry variables retained in multivariable stepwise regression models predicting each calcification measure in the pooled FHS Third Generation and Offspring Cohorts without prevalent CVD (n=2920)

	OR (95% CI)	P value
Prevalent TAC
CFPWV	2.77 (2.21-3.47)	<0.0001
CPP	1.43 (1.19-1.72)	0.0002
Prevalent AAC
CFPWV	1.41 (1.19-1.66)	<0.0001
CPP	1.29 (1.09-1.51)	0.0027
AI	1.37 (1.19-1.57)	<0.0001
Prevalent CAC
CFPWV	1.51 (1.29-1.76)	<0.0001

AI: augmentation index. CFPWV: carotid-femoral pulse wave velocity. CAC: coronary artery calcium. CPP: central pulse pressure. TAC: thoracic aortic calcium. AAC: abdominal aortic calcium.

Multivariable models adjusted for age, sex, cohort, body mass index, height, heart rate, history of hypertension or lipid-lowering medications, total cholesterol/HDL, diabetes, and current smoking.

OR represent odds for outcome of prevalent calcification per SD increment in tonometry measure.

Forward wave amplitude was not associated with any calcification measure in MV models containing the other tonometry measures.

We also evaluated the association of augmentation pressure with AAC and found similar relations to that between AI and AAC in multivariable and SBP-adjusted models (B±SE 0.60±0.09 and OR for prevalent AAC 1.49, 95% CI 1.30-1.72, p<0.001) In addition, to further examine the association of AI with AAC, we evaluated the relations of wave reflection timing, reflected wave amplitude, and reflection coefficient with AAC. The timing of wave reflection was inversely associated with both log_e(AAC+1) (B±SE = −0.26±0.08, p<0.001) and prevalent AAC (OR 0.84, 95% CI 0.75-0.93, p<0.001) in multivariable adjusted models that included SBP. Similar results were obtained in a logistic regression model that defined calcification by the 90^th percentile criterion (OR 0.79, 95% CI 0.70-0.90, p<0.001). Reflected wave amplitude was associated with log_e(AAC+1) (B±SE = 0.31±0.11, p=0.003) and with prevalent AAC (OR 1.29 (1.12-1.49), p<0.001) in multivariable models including SBP. In contrast, reflection coefficient, the ratio of reflected wave and forward wave amplitude, was not significantly associated with either continuous (B±SE= 0.18±0.09, p=0.05) or prevalent AAC (OR 1.08, 95% CI 0.96-1.22, p=0.19).

DISCUSSION

In our large cross-sectional study of ambulatory adults, we observed strong and variable associations of arterial stiffness, pressure pulsatility, and wave reflection with the presence and extent of arterial calcifications. We also observed cohort differences in associations of arterial stiffness and AI, the end effect of greater aortic stiffness and attenuation in wave reflection after middle age.²³ Stepwise models showed that CFPWV and CPP, a measure of pressure pulsatility, were jointly associated with TAC and AAC, suggesting their associations with aortic calcification and hemodynamic load. Additionally, CFPWV, a measure of global aortic stiffness, was associated with coronary calcification. Measures of wave generation and reflection were associated with calcification of different regions of the aorta, with forward wave amplitude associated with TAC and both AI and reflected wave amplitude associated with AAC. Moreover, the relations between arterial stiffness and subclinical calcification had larger effect sizes in the younger rather than older individuals, suggesting the pathophysiology between the two is present by mid-adulthood.

Arterial stiffness and vascular calcification share common CVD risk factors^{3, 6, 8, 10} The pathophysiology of arterial stiffening involves fragmentation of elastin fibers in the arterial wall, extracellular matrix remodeling and deposition of collagen, which is markedly stiffer than elastin.^{22, 24} Elastin degradation, as may occur from adverse central hemodynamics, leads to medial calcium deposition in animal models²⁵ and medial calcification may lead to further destruction of elastin.¹³ While our CT technique, due to the lack of spatial resolution needed, cannot differentiate between intimal and medial coronary artery calcification, histopathology studies have shown that the overwhelming majority of calcification in subjects without known renal disease can be found in the intima of the coronary arteries.²⁶ Thus, medial calcification is unlikely to explain the associations observed in our study. However, it is possible that intimal and medial calcification occur in parallel, such that intimal calcification by MDCT is a marker of medial calcification.

Greater stiffness of the aortic wall leads to a higher forward wave amplitude and propagation velocity. Consistent with CFPWV and CPP as measures of global stiffness and pressure pulsatility along the aorta, respectively, we found strong relations of each with TAC and AAC after accounting for coexistent risk factors. The stronger relations of CFPWV and CPP with TAC, rather than AAC, highlight the dominant influence of proximal, as compared with distal, aortic structure and function to global stiffness measures. However, distal arterial properties likely still influence proximal hemodynamics. Wave reflection in the arterial system arises in regions of impedance mismatch, such as a narrowing, branch point or change in wall properties.²⁷ The association of AI and an earlier timing of wave reflection with AAC suggests that calcification of the abdominal aorta may create a discrete reflecting site that enhances the effects of wave reflection on proximal aortic pressure.

The relations of adverse arterial hemodynamics and arterial calcification may be bi-directional. Elevated aortic stiffness and wave reflection may alter local flow and pressure in the aorta, creating a milieu that facilitates atherogenesis and calcification. Alternatively, intimal calcification and resultant narrowing of the vessel caliber may create abnormal arterial hemodynamics. The association of arterial stiffness and CAC is of interest, particularly as the association persists in the absence of aortic calcification. Arterial stiffness may result in elevated coronary perfusion pressure during systole and lower than usual flows during diastole, potentially resulting in endothelial injury and atherosclerotic plaque formation. Severe atherosclerosis and resting obstruction to coronary flow may lead to left ventricular systolic and/or diastolic dysfunction, in turn resulting in abnormal flow hemodynamics and global aortic stiffness. However, the lack of prevalent CVD in our sample, and finding of a stronger association of aortic stiffness with CAC in the younger cohort with a more favorable CVD risk factor profile, both argue against the latter possibility. Additional longitudinal studies, particularly beginning at younger ages, may be able to discern the temporal nature of these events.

CFPWV was associated with aortic calcification in a cohort of healthy older people.¹⁶ In that study, CPP and AI were not associated with calcification in adjusted models, which may be due to the smaller sample size or related to the extent and distribution of calcification. Our results are generally consistent with other large community-based studies that have examined the association of local properties of the aorta with TAC and AAC. The Multi-Ethnic Study of Atherosclerosis (MESA) reported that common carotid artery distensibility, an indirect measure of the stiffness of the aorta, was associated with TAC prevalence and score, with a stronger unadjusted correlation with TAC than with AAC.¹⁴ A subsequent analysis in the MESA cohort found an association between local ascending aortic distensibility and TAC.¹⁹ In contrast to MESA, we evaluated measures of global aortic function, such as CFPWV and CPP, which have been related to major adverse clinical events.^2-5 In addition, we evaluated relations of wave reflection and aortic calcification, which have not been well-described in large epidemiological samples.

Moreover, we observed greater effect estimates in the continuous relations between arterial stiffness and calcification among the younger aged cohort, as compared with the older cohort. Thus, our findings support a strong connection of adverse arterial hemodynamics and subclinical atherosclerosis early in life, and further emphasize the need for further investigation into the potential benefit of early identification and treatment on preventing or slowing disease progression.

Prior reports have examined components of arterial stiffness in isolation with AAC. To our knowledge, a comprehensive examination of the relations of arterial stiffness, pulsatility, and wave reflection to AAC has not reported in a large community-based cohort without prevalent diagnosed CVD. CFPWV has been related to AAC, as assessed by the prevalence of calcification in the abdominal aorta identified by lumbar radiographs in the Rotterdam study²⁸ and by CT scan in a multiethnic cohort of younger adult men¹⁸. Our study expands the breadth of prior work by providing a simultaneous assessment of the association of various measures of aortic stiffness and pressure pulsatility, including CFPWV, CPP, forward wave amplitude, and AI with regional measures of aortic calcification along the full length of the aorta and in the coronary arteries.

Two groups have reported relations between aortic stiffness and CAC in community-dwelling adult cohorts.^{20, 29} These studies showed that greater CFPWV was associated with the presence and extent of CAC. Extending these investigations, we examined the relations of other measures of pressure pulsatility and wave reflection with CAC, and found that CFPWV had the strongest association with CAC. Furthermore, the association of CFPWV with CAC in our sample persisted in a model that included only cases without aortic calcification, suggesting that the association represents something other than a “bystander” effect of pan-arterial calcification.

Our study is cross-sectional; therefore, we are not able to infer causality or determine the directionality of effects. Relations between hemodynamics and calcification are likely bidirectional and may facilitate a feedback loop in which arterial stiffness begets calcification, and vice versa. Additional longitudinal studies may be able to discern the temporal nature of these events. Our study sample contained many participants with treated hypertension. As blood pressure lowering may affect the relation of tonometry measures to arterial calcification, a study of untreated or poorly controlled individuals with hypertension may further elucidate the relations between arterial stiffness, pressure pulsatility, and wave reflection with arterial calcification. Additionally, the FHS cohorts are predominantly of white European ancestry, and thus our findings may not be fully generalizable to other ethnicities. However, our results are consistent with prior published reports relating arterial stiffness to calcification in multi-ethnic groups.^{14, 18, 19} We evaluated 4 tonometry phenotypes and 3 calcification phenotypes and did not adjust for multiple testing. Because of correlations within the tonometry and calcification phenotypes, a strict Bonferroni adjustment would be overly conservative and may be associated with false negatives. Importantly, many of our associations were extremely robust; less robust associations will require independent confirmation. Advantages of our study include routinely ascertained phenotypes in a well-characterized community-based sample free of overt CVD spanning a broad adult age range.

In conclusion, in our community-based cohort without clinically manifest CVD, we demonstrated strong associations of arterial stiffness, pressure pulsatility, and wave reflection with multi-territory arterial calcification. Moreover, the observed associations, which persisted even in our youngest participants, support further investigation into the benefit of intervention at an early subclinical stage in order to attenuate or prevent progressive stiffening and calcification with advancing age. Further prospective studies will be required to examine temporal relations of arterial stiffening and calcification in order to better understand mechanisms contributing to the associations.

SIGNIFICANCE.

Both arterial stiffness and calcification are sources of comorbidity and mortality associated with aging and CVD risk. Yet, their relations remain incompletely understood. We comprehensively examined the associations of aortic stiffness, pressure pulsatility, and wave reflection with both aortic and coronary calcification. Aortic stiffness and pressure pulsatility were associated with predominantly thoracic, but also abdominal aortic calcification in a community cohort without prevalent CVD. Aortic stiffness was additionally associated with coronary calcification even in participants with no aortic calcification. Despite a lower prevalence of cardiovascular risk factors in younger individuals, there was a stronger association between arterial hemodynamics and calcification in this group. Our results suggest different mechanisms by which abnormal arterial hemodynamics may beget atherosclerosis and calcification. These associations are present at earlier ages than previously recognized and suggest a need for further investigation into the temporal development of preclinical abnormal hemodynamics and atherosclerosis at younger ages.

ABBREVIATIONS AND ACRONYMS

AAC

Abdominal aortic calcification

AI

Augmentation index

CAC

Coronary artery calcification

CFPWV

Carotid-femoral pulse wave velocity

CPP

Central pulse pressure

CVD

Cardiovascular disease

FHS

Framingham Heart Study

MDCT

Multi-detector computed tomography

OR

Odds ratio

SBP

Systolic blood pressure

TAC

Thoracic aortic calcification