Association of air pollution with longitudinal changes in arterial stiffness and correlates of longitudinal change in arterial stiffness in the Multi-Ethnic Study of Atherosclerosis (MESA)

BACKGROUND

Many studies have shown associations between particulate matter with diameter of less than 2.5 micrometers in aerodynamic diameter (PM2.5), also called fine particulate air matter, and cardiovascular disease (CVD) events. Improved understanding of the biological mechanisms linking air pollution to cardiovascular health effects is crucial to giving further support and justification for limiting air pollution. Arterial function measures, which measure arterial elasticity and stiffness, may be part of mechanisms linking air pollution and CVD events. Limited research has been done to examine how predictive a wide range of arterial stiffness measures is of CVD events in a single cohort. Few studies have examined determinants of longitudinal change in arterial stiffness.

OBJECTIVES

The overall objective of this dissertation is to examine whether long-term air pollutant exposures impact cross-sectional measurements of arterial stiffness and longitudinal change in arterial function measures. We also investigated how short-term air pollutant exposures affect cross-sectional measurements of arterial function. We looked at how these arterial stiffness measures predict subsequent cardiovascular disease (CVD) events. In addition, the correlates of longitudinal change in arterial stiffness measures were also examined in order to better understand how these measures may be on the biological pathway connecting traditional cardiovascular risk factors and CVD events.

METHODS

First, we examined compared the predictive value of five measures of arterial function, both arterial stiffness and arterial elasticity, assessed at Exam 1 of the Multi-Ethnic Study of Atherosclerosis (MESA study) (2000-2002) on the time to occurrence of the first coronary heart disease event (CHD) in the MESA study using Cox proportional hazards regression. To assess the ability of each arterial function measure to improve discrimination of events, we used receiver-operating characteristic curves and areas under the receiver-operating characteristic curves (AUC). The arterial function measures we used for this analysis included C1 and C2, which are measures of arterial elasticity obtained via radial tonometry, aortic distensibility (AD), which is a measure of arterial elasticity obtained via aortic MRI, and carotid distensibility (CD) and Young’s modulus (YM) at the carotid artery, which are measures of arterial elasticity and arterial stiffness obtained via carotid ultrasound. Next, we investigated the association between traditional cardiovascular risk factors and longitudinal change in four arterial functional measures between Exam 1 (2000-2002) and Exam 5 (2010-2012) of the MESA study utilizing linear mixed models with a fixed slope and random intercepts. As sensitivity analyses, we also examined these relationships using simple linear regression models. Finally, we utilized linear mixed models to examine the association between long-term air pollution and cross-sectional measurements of arterial function, long-term air pollution and longitudinal change in arterial function measurements, and short-term pollution and cross-sectional measurements in arterial function measurements. The Multi-Ethnic Study of Atherosclerosis Air study (MESA Air) estimated the outdoor ambient individual concentrations of fine particulate matter (PM2.5) and oxides of nitrogen (NOx) based on spatiotemporal air pollution exposure models. These models incorporated a wide range of geographic covariates and data from cohort-specific air pollution models. For short-term air pollution analyses, we utilized city-wide daily average fine particulate matter (PM2.5) from Air Quality System (AQS) monitors. These arterial function measures used to investigate the relationship of traditional cardiovascular risk factors with arterial function measures and the relationship between air pollution exposure with arterial function measures included Pressure Time Constant 1 (PTC1) and Pressure Time Constant 2 (PTC2), which are measures of arterial elasticity obtained via radial tonometry, and distensibility coefficient (DC) and Young’s Elastic modulus (YEM) at the carotid artery, which are measures of arterial elasticity and arterial stiffness obtained via carotid ultrasound. PTC1 and PTC2, as well as DC and YEM, are calculated using slightly different formulas than C1 and C2, as well as CD and YM, respectively.

RESULTS

In our analysis of the predictive value of arterial function measured at baseline of the MESA study, we found that the hazard ratio of CHD event per standard-deviation higher value of arterial function was 0.97 (95% Confidence Interval (CI): 0.86, 1.10) for C1, 0.73 (95% CI: 0.63, 0.86) for C2, 0.98 (95% CI: 0.86, 1.11) for carotid distensibility, 0.99 (95% CI: 0.90, 1.09) for Young’s modulus, and 0.90 (95% CI: 0.74, 1.10) for aortic distensibility. C2 provided additional discrimination for the prediction of CHD (area under the curve= 0.736 vs. 0.743, p=0.04). Arterial stiffness increased in all measures as anticipated over time, though different risk factors were associated with changes in specific arterial function measures: Being of male gender was associated with larger declines for PTC2. Higher heart rate was associated with smaller declines in PTC1. Higher BMI at baseline was associated with smaller declines in DC. Having diabetes at baseline was associated with larger increases in YEM. Increased age at baseline was associated with smaller declines in PTC1, PTC2, DC (deceleration of stiffening), but larger increases in YEM (acceleration of stiffening) over follow-up. In addition, increased systolic arterial pressure (SBP) at baseline and as a time-varying variable were associated with smaller declines in PTC1, PTC2. However, higher mean arterial pressure (MAP) was associated with larger increases in YEM (acceleration of stiffening). We found that a one interquartile (IQR) increase in annual average exposure to NOx during year 2000 (IQR= 44.5 ppb) was associated with a 0.220 (seconds*10)-1 decrease in PTC1 (95% confidence interval (CI): -0.437 to -0.003) measured at Exams 1 and 5 (less arterial elasticity). We also found associations between increased short-term air pollution exposure with smaller cross-sectional measurements of arterial elasticity (less elasticity) and larger cross-sectional measurements of arterial stiffness (more stiffness). There were no statistically significant associations between long-term pollutant exposures and the rate of change in any of the arterial function measures.

CONCLUSIONS

Our findings provide additional data on the relationships between CVD risk factors, arterial functional measurements and CVD events. Of the five arterial function measures at baseline that we analyzed, only C2 showed an association with subsequent CHD events. Lower C2 at baseline was associated with higher risk of future CHD events over follow-up. We observed different directions of the association of age with longitudinal change in arterial stiffness and between blood pressure measures with longitudinal change in arterial stiffness. We observed that higher mean arterial pressure and having increased age at baseline were associated with larger increases in YEM over follow-up. However, we saw higher systolic blood pressure was associated with smaller declines in PTC1 and PTC2. Also, increased age at baseline was associated with smaller declines in PTC1, PTC2, and DC. Higher long-term traffic-related air pollution was associated with less arterial elasticity, when assessed by cross-sectional measurement of PTC1 at Exam 1. Higher short-term fine particulate air matter exposure was associated with reduced arterial elasticity and increased arterial stiffness, as assessed by PTC2, DC, and YEM. However, these findings do not support that long-term air pollution is associated with rate of change in arterial function measures. Additional studies of the short-term effects of air pollution and arterial function measures may help us better understand the biological mechanisms by which air pollution is related to cardiovascular disease events.