|Home | About | Journals | Submit | Contact Us | Français|
Exposure to ambient air pollutants increases risk for cardiovascular health outcomes in adults. The contribution of childhood air pollutant exposure to cardiovascular health has not been thoroughly evaluated.
The Testing Responses on Youth study consists of 861 college students recruited from the University of Southern California in 2007–2009. Participants attended one study visit during which blood pressure, heart rate and carotid artery intima-media thickness (CIMT) were assessed. Self-administered questionnaires collected information about health and socio-demographic characteristics and a 12-hr fasting blood sample was drawn for lipid and biomarker analyses. Residential addresses were geocoded and used to assign cumulative air pollutant exposure estimates based on data derived from the U.S. Environmental Protection Agency’s Air Quality System (AQS) database. The associations between CIMT and air pollutants were assessed using linear regression analysis. Mean CIMT was 603 μm (± 54 SD). A 2 standard deviation (SD) increase in childhood (aged 0–5 years) or elementary school (aged 6–12) O3 exposure was associated with a 7.8 μm (95% CI −0.3, 15.9) or 10.1 μm (95% CI 1.8, 18.5) higher CIMT, respectively. Lifetime exposure to O3 showed similar but non-significant associations. No associations were observed for PM2.5, PM10 or NO2 although adjustment for these pollutants strengthened the childhood O3 associations.
Childhood exposure to O3 may be a novel risk factor for CIMT in a healthy population of college students. Regulation of air pollutants and efforts that focus on limiting childhood exposures continue to be important public health goals.
Cardiovascular diseases (CVD) remain a leading cause of morbidity and mortality.1 Given the heavy disease burden CVD presents, it is important to identify early life risk factors of CVD in order to target prevention strategies and reduce lifetime risk. Assessment of carotid artery intima-media thickness (CIMT), a validated and reliable marker of atherogenesis,2 in young adults provides a promising approach for investigating early life risk factors for adult clinical cardiovascular outcomes.
Ample evidence suggests that exposure to ambient air pollutants increases risk for cardiovascular health outcomes.3 Exposure to air pollutants has been related to acute CVD events including myocardial infarction, stroke, and mortality.4–7 While animal studies provide strong evidence for an atherogenic role of ambient air pollutants, namely particulate matter (PM), only a few studies have investigated this hypothesis in humans. Particulate matter and traffic-related pollutants have specifically been associated with level and progression of CIMT in adults.8–10 However, atherogenesis is a lifelong process with fetal and early postnatal origins.11 Currently it is not clear how early in life pollution may begin to exhibit its atherogenic effects or whether childhood exposures to air pollutants affect later cardiovascular health.
Exposure to high levels of ambient air pollutants in childhood may contribute to subtle yet detectable changes in cardiovascular health, thereby predisposing these children to earlier development of cardiovascular pathologies and disease later in life. In the few studies conducted in healthy populations of children or young adults, childhood or recent exposure to air pollutants has been associated with increases in blood pressure and heart rate12 and with arterial stiffness but not with CIMT.13–14 Moreover, pollutants such as PM2.5 have been associated with systemic inflammation, oxidative stress, insulin resistance, and endothelial injury in children and young adults.15–17
We designed a study to investigate whether early childhood exposure to air pollutants has consequences on cardiovascular health indicators in young adults. In the current report, we investigated the association between childhood and lifetime cumulative average exposures to PM10, PM2.5, NO2 and O3 with CIMT in a population of University of Southern California college students. Differences in susceptibility to exposure by sex, body mass index (BMI), ethnicity and playing outdoor team sports were also investigated.
The Testing Responses on Youth (TROY) study consists of 861 college students recruited from USC in 2007–2009. The primary purpose of the TROY study is to assess lifetime histories of air pollution exposure in relation to early determinants of atherosclerosis. Participants were eligible for study inclusion if they were lifetime non-tobacco smokers, were born in the United States or moved to the United States within the first six months of life, attended high school in a large city in the United States, and provided written informed consent to participate.
Participants attended a study visit during which CIMT, systolic and diastolic blood pressure, heart rate, height, weight, and lung function were measured. CIMT, heart rate, and blood pressure were assessed by a single physician-imaging specialist from the USC Atherosclerosis Research Unit Core Imaging and Reading Center. Several self-administered questionnaires were completed during or prior to the office visit to gather information about health and socio-demographic characteristics (see online data supplement p.1).
Participants provided a 12-hr fasting blood sample for lipid and biomarker analyses following completion of health testing. Of the 861 initially screened participants, 16 were excluded for not meeting eligibility criteria and 71 did not provide blood samples or were missing lipid or biomarker measurements, and 6 were excluded because they were born and lived outside of the US for more than 6 months after their birth, leaving 768 participants in the study population.
The study protocol was approved by the institutional review board for human studies at the University of Southern California, and written consent was provided by the study subjects.
High-resolution B-mode ultrasound images of the right common carotid artery (CCA) were obtained with a portable Biosound MyLab 25 ultrasound system attached to a 10-MHz linear array transducer and read by a single physician-imaging specialist. As described previously described (Patents 2005, 2006, 2011)18–19, the jugular vein and carotid artery were imaged transversely with the jugular vein stacked above the carotid artery. All images contained internal anatomical landmarks for reproducing probe angulation and a three lead electrocardiogram was recorded simultaneously with the B-mode image to ensure that CIMT was measured at the R-wave in the cardiac cycle. CCA far wall IMT was determined as the average of 70 to 100 measurements between the intima-lumen and media-adventitia interfaces along a one-cm length just proximal to the carotid artery bulb by automated computerized edge detection with an in-house developed software package (Patents 2005, 2006, 2011).18–19 This method standardizes the timing, location, and distance over which CIMT is measured, ensuring comparability across participants (see online data supplement p. 1).18–19
Blood pressure and heart rate were measured immediately after the ultrasound examination by standard techniques after the subject was recumbent for at least ten minutes. Blood pressure was measured three times in one-minute intervals, using an OMRON blood pressure monitor with automatic cuff inflation and deflation. Heart rate was measured using a three lead electrocardiogram as part of the Biosound MyLab 25 ultrasound system. Subject standing height was measured in stocking feet to the nearest centimeter using a metal measuring tape placed perpendicularly to the floor through the use of a construction-type bubble level and a measurement block to properly align head orientation. Weight was measured to the nearest pound with a medical-grade scale calibrated prior to each day’s testing using pre-determined calibration weights.
Plasma and serum were divided into one ml samples and stored at −80 degrees Celsius until analyzed. One ml of plasma from each subject was used to measure total cholesterol, triglyceride, and HDL cholesterol levels using an enzymatic method in conformance with the Standardization Program of the National Centers for Disease Control and Prevention. LDL-C was calculated using the Friedwald formula.18
One ml of serum from each subject was used to measure CRP. High-sensitivity CRP was measured by a solid-phase chemiluminescent immunometric assay using the Immulite 2000 analyzer (Siemens Medical Solutions Diagnostics, Malvern, PA). The sensitivity of the assay was 0.02 mg/dL and the inter-assay coefficient of variation was 7.0%.
A detailed lifetime residential history was completed by participants. Participant residence addresses within the U.S. were standardized and their locations were geocoded using the Tele Atlas Geocoding Service (Tele Atlas Inc., Menlo Park, California, www.na.teleatlas.com). Of the 2,598 residential locations reported, 98.3% (2,553) were U.S. residences that were successfully geocoded. Of the 2,553 participant residence locations: 47.8% were geocoded with the highest quality match, which is specific to the centroid of the parcel or building footprint (Address Point match), and 46% were geocoded to the street segment and/or the relative position between nearest intersections. The remaining 6.2% were geocoding using Google Earth. No assignments were made for the 45 participant locations that could not be geocoded.
Ambient air pollution concentrations were estimated for each subject’s residence within the U.S. from the time the subject occupied that residence to the participant’s CIMT measurement. Move-in and move-out dates were provided for each residence, and ambient air quality data was spatially interpolated to those locations for the relevant time periods. The station-specific monthly air quality data were spatially interpolated using inverse distance-squared weighting (IDW2). The data from up to four air quality measurement stations were included in each interpolation. Due to the regional nature of O3, NO2, PM10, and PM2.5 concentrations, a maximum interpolation radius of 50 km was used for all pollutants. However, when a residence was located within 5 km of one or more stations with valid observations, the interpolation was based solely on the nearby values.
Early childhood exposure corresponding to the elementary school years (separately for ages 0–5 and 6–12) and lifetime exposure (from birth to date of CIMT measurement) were calculated by averaging exposures across the relevant residential histories for those time periods. In the event that participants had multiple concurrent residences during the exposure periods of interest, time was split by weighting the location-specific exposure estimates according to the appropriate number of summer or non-summer days per week at that location.
Air pollutant estimates were derived from the U.S. Environmental Protection Agency’s Air Quality System (AQS) database for the years 1980 through 2009. Hourly concentrations of O3 and NO2, and daily concentrations of PM10, PM2.5 measured in all 50 states for January 1980 through 2009 were downloaded from AQS. The PM data were primarily limited to those collected with Federal Reference Method (FRM) monitors and Federal Equivalent Method (FEM). Non-FEM PM2.5 data were used when no FEM measurements were available. Automated quality control checks on the concentration ranges and persistence were applied to the AQS data. Few PM10 and PM2.5 data exist prior to 1987 and 1999, respectively. The AQS data were augmented in southern California with O3, NO2, PM10, and PM2.5 data from the Children’s Health Study (CHS) for 1994–2009.20–21 National-scale PM10 data were filled in using adjusted total suspended particulates (TSP) data for 1981–1987. Pre-1999 PM2.5 data for southern California were filled in with 1994–1998 estimated PM2.5 concentrations developed for the CHS. In order to assign a childhood or lifetime exposure estimate, data were required to be 75% complete for O3 and NO2 and 12% for PM to account for the one-in-six day sampling. As a result, of 768 initial study subjects, 12 were missing PM10 values, 27 were missing PM2.5, 115 were missing NO2, and 71 were missing O3 for childhood exposure and 6 were missing PM10 values, 20 were missing PM2.5, 75 were missing NO2, and 91 were missing O3 for elementary exposure. More details are provided in the online data supplement (p. 1).
The associations between CIMT and air pollutants were assessed using linear regression analysis. Variables initially evaluated for confounding based on whether they changed the effect estimate by greater than 10% and subsequently dropped for lack of evidence included family history of cardiovascular disease, current physical activity, participation in outdoor team sports during childhood, and mother’s education. A final multivariate model adjusted for age, sex, race/ethnicity, BMI and systolic blood pressure, second hand smoke in childhood, current second hand smoke, hsCRP, LDL-C, and HDL-C.
To examine whether the associations between air pollutants and CIMT varied by sex, BMI, race/ethnicity, or participation in team sports, we included interaction terms in the regression models and used likelihood ratio tests to evaluate overall significance of the interactions. We conducted a series of sensitivity analyses to evaluate whether exclusion of 37 participants who reported a family history of heart disease or 321 participants who reported a family history of hypertension or high cholesterol affected our results. We also evaluated whether blood pressure or lipid levels might mediate in the association between O3 exposures and CIMT by comparing effect estimates from two models, one with and one without each of these variables included. Lastly, we restricted the analysis to participants from southern California on whom we had supplemental air monitoring data and we restricted to participants with available data for all the different pollution variables in order to compare results among participants with similar exposure assessment. We also evaluated whether data quality affected results by restricting to participants with the best quality data for all time periods. Regression procedures were conducted in SAS.22 All statistical testing was conducted with a two-sided alpha level of 0.05.
Baseline characteristics of the 768 study participants are shown in Tables 1 and and2.2. Little difference was observed between study participants and the 83 students who were excluded because of missing data or eligibility. All participants were college students who were on average 20 years of age; the sample included more females (59%) than males (41%). Only one participant had high blood pressure (defined as > 120/80 mmHg) and family history of heart disease (5.5%) was rare in this population.
CIMT measurements were highly reproducible. The coefficient of variation between replicate scans (n=93) was 0.83% and the correlation coefficient was 0.98. CIMT was normally distributed in this population, with a mean (SD) of 603.4 μm (54.5 μm), as described previously.23
Participants’ residences were widely distributed across the country (Figure 1, online data supplement Table 1) and air pollutants had a broad range of distribution across various time periods in early childhood and lifetime cumulative averages (online data supplement Table 2). In general, cumulative averages for the same pollutant across different time periods were highly correlated with one another and PM2.5, PM10 and NO2 were highly correlated with one another within each time period (online data supplement Table 3). Ozone was not highly correlated with the other pollutants.
Ozone exposure during elementary school years (ages 6–12) was associated with a 10.1 μm (95% CI 1.8, 18.5) higher CIMT per 2 standard deviation (9.3 ppb) difference in O3 (Table 3). Early childhood exposure (between 0–5 years) and lifetime exposure to O3 showed similar but non-significant associations. No associations between CIMT and NO2, PM10 and PM2.5 were observed. Adjustment for multi-pollutant models in which O3 and a second pollutant (either NO2, PM10 or PM2.5) were simultaneously modeled in many cases strengthened the association with O3, particularly in early childhood (ages 0–5) (Table 4). The effects on CIMT of O3 exposure during early childhood (ages 0–5) ranged from 8.5 to 10.0 μm depending on the additional pollutant adjusted in the model.
Further investigations into whether the association between air pollutants and CIMT varied by sex, ethnicity, playing on outdoor team sports, or overweight/obesity showed no significant interactions. Moreover, we found no evidence for synergistic effects between multiple pollutants on CIMT.
We also evaluated whether blood pressure or lipid levels might mediate the association between O3 exposures and CIMT. Inclusion or exclusion of SBP, DBP, HDL-C or LDL-C from the models did not substantially change the effect of O3 (data not shown), suggesting that these measures are not important mediators of the observed pollutant effects. Sensitivity analyses were conducted to evaluate whether exclusion of 37 participants who reported a family history of heart disease or 321 participants who reported a family history of hypertension or high cholesterol affected our results but no effects were observed (online data supplement Table 4). We also conducted the following sensitivity analyses: 1) we restricted the analysis to participants from southern California on whom we had supplemental air monitoring data (online data supplement Table 5); 2) we restricted to participants with available data for all the different pollution variables (online data supplement Table 6); and 3) we restricted to participants with the best quality data for all time periods defined as having at least one monitor within 50 miles of the address (online data supplement Table 7). In each case the results were similar to our main findings for ozone.
Childhood exposure to O3 was associated with increased CIMT in young adulthood. We observed that the effects of O3 exposure were evident in young adults, even in a select group of healthy, non-smoking college students who may not be representative of the general population.
Studies in animal models have identified a biological mechanism for O3 effects on cardiovascular health. Chronic exposure to high O3 resulted in 179% thicker intima and media of peribronchiolar arterioles in a monkey model 24 and O3 significantly modulated vascular tone regulation and increased oxidant stress and mitochondrial DNA damage in a mouse model.25 In these apoE−/− mice, O3 exposure resulted in significantly increased atherogenesis compared with filtered air controls.25 These experiments suggest that O3 – induced tissue injury that begins in the lung can be further propagated by endogenously produced generation of reactive oxygen species and oxides of nitrogen.26 Specifically, Cole et al suggest that primary O3 reactions could oxidize key inflammatory-related transcription factors, thereby triggering a cascading pro-inflammatory state outside of the lung that could initiate and propagate inflammatory responses in the vascular compartment.26
Chronic O3 exposure in adults has been associated with respiratory mortality 27 and early life exposure to O3 has been associated with childhood lung function development. 28–30 Several studies have also suggested links between lung function deficits and cardiovascular risk 31–33, including associations between reduced lung function and CIMT.34 Such population-based studies support the hypothesis suggested in animal models that systemic inflammation induced first in the lung may be propagated throughout the body to target other organ systems.26
In humans, cholesterol ozonolysis is present in atherosclerotic tissue and its byproducts are thought to promote the pathogenesis of atherosclerosis.35 In vitro studies of human blood cells have shown that O3 increases lipid peroxidation, protein thiol content, systemic oxidative stress and induction of some cytokines.36 In population-based studies, O3 exposure has been associated with cardiovascular-related hospital admissions and mortality.27, 36 Exposure to concentrated ambient particles plus O3 in adult volunteers caused brachial artery vasoconstriction compared with filtered air inhalation.37 In a small occupational cohort of mail carriers among whom personal O3 and size-fractionated particulate matter exposures were monitored, O3 was associated with increased ankle-brachial index but not with heart rate variability.38 However, in both studies the individuals evaluated were older adults and the exposures were acute.
While there is biological support for the involvement of O3 in atherosclerosis, evidence that long-term O3 exposure independent from particulate matter exposure affects cardiovascular health outcomes in epidemiologic studies remains inconclusive. Even less is known about susceptibility to O3 exposure during childhood. In this study, we evaluated the effects of childhood and lifetime O3 exposure on cardiovascular health in a large group of young adults. We further investigated multi-pollutant models in which we included O3 and particulate matter jointly and found the O3 results were robust and independent of any associations with particulate matter. We provide some of the first evidence to suggest that childhood O3 exposure is associated with elevated subclinical atherosclerosis, measured by CIMT in college students averaging 20 years of age.
We did not observe any associations with PM2.5, PM10 or NO2 and CIMT. The lack of association with PM is consistent with a Dutch study based on young adults 13 but inconsistent with studies done in older ages.6 The biological mechanisms through which different air pollutants act on the vasculature may differ and longer periods of time may be necessary to observe cardiovascular health effects of cumulative PM exposures than O3, particularly for anatomical changes such as CIMT. At present, too few studies evaluating childhood exposures exist to draw meaningful conclusions on the relative importance of O3 compared to particulate exposures with respect to cardiovascular risk later in life. Although Lenters et al did not observe an association with CIMT, they did observe changes in arterial stiffness in response to pollutant exposures.13 In future studies in the TROY population we will evaluate the potential associations between air pollutants and arterial stiffness, a physiological change that may respond to exposure more quickly.
One of the strengths of this large cross-sectional study is the availability of cumulative lifetime air pollutant exposure histories for participants. However, given the high degree of correlation between early childhood (ages 0–5), elementary school years (ages 6–12) and lifetime exposure estimates (birth - CIMT test date), we cannot conclude with certainty that one time period is a more critical window of susceptibility than another. We also evaluated whether the observed effects of O3 might be mediated through effects on blood pressure or lipid levels but found that inclusion or exclusion of these terms did not alter the associations.
Because we calculated air pollutant exposure estimates over an 18-year time frame, measurement error may be of some concern. The quality of air pollutant data varied over time and space, for example. However, sensitivity analyses restricting the dataset to participants with only the highest quality data yielded similar results. In addition, lack of an association between particles and CIMT may be due to increased measurement error and lack of power in these metrics, particularly for PM2.5. A lack of monitoring data for PM2.5 in early years resulted in a smaller sample size for those analyses. In addition, imputation of PM2.5 values based on historical PM10/PM2.5 ratios may have increased the level of error. Alternatively, the observed association with O3 but not PM could be due to different pathophysiology. That is, perhaps O3 is an important risk for atherogenesis whereas PM is important in atherothrombosis or plaque rupture. Nevertheless, for ozone we observed a consistency in effect estimates across various time points, and in several sensitivity analyses, further supporting an association despite potential differences in measurement error.
In conclusion, the atherogenic process has important determinants early in life. We present evidence that childhood exposure to O3 is a novel risk factor for CIMT in a healthy population of college students. Regulation of air pollutants and efforts that focus on limiting childhood exposures continue to be important public health goals to potentially reduce atherosclerosis burden and its consequences.
The atherogenic process has important determinants early in life. Childhood exposure to ozone may be a novel risk factor for CIMT in young adults. Exposure to ambient air pollutants increases risk for cardiovascular health outcomes in adults. Early life exposure to ozone has been associated with childhood lung function development. Systemic inflammation induced first in the lung may be propagated throughout the body to target other organ systems. Regulation of air pollutants and efforts that focus on limiting childhood exposures continue to be important public health goals to potentially reduce atherosclerotic burden and its consequences.
This work was supported by the Southern California Environmental Health Sciences Center (grant # 5P30ES007048) funded by the National Institute of Environmental Health Sciences and NIEHS grants 5R01ES014708, 5P30ES007048, 1K01ES017801, 5P01ES009581, R826708-01 and RD831861-01.
No conflicts of interest declared.