Skip to main content
Download PDF
Download ePub

Associations of time-weighted individual exposure to ambient particulate matter with carotid atherosclerosis in Beijing, China

Environmental Health volume 22, Article number: 45 (2023) Cite this article

Abstract

Background

Time-location information (time spent on commuting, indoors and outdoors around residential and work places and physical activity) and infiltrated outdoor pollution was less considered estimating individual exposure to ambient air pollution. Studies investigating the association between individual exposure to particulate matter (PM) with aerodynamic diameter < 10 μm (PM10) and < 2.5 μm (PM2.5) and carotid atherosclerosis presented inconsistent results. Moreover, combined effect of pollutants on carotid atherosclerosis was not fully explored. We aimed to investigate the association between long-term individual time-weighted average exposure to PM2.5 and PM10 and the risk of carotid atherosclerosis, and further explore the overall effect of co-exposure to pollutants on carotid atherosclerosis.

Methods

The study population included 3069 participants derived from the Beijing Health Management Cohort (BHMC) study. Daily concentration of ambient air pollutants was estimated by land-use regression model at both residential and work addresses, and one- and two-year time-weighted average individual exposure was calculated by further considering personal activity pattern and infiltration of ambient air pollution indoors. We explored the association of PM2.5 and PM10 with carotid atherosclerosis and pooled the overall effect of co-exposure to ambient air pollutants by quantile g-computation.

Results

A significant association between time-weighted average exposure to PM2.5 and PM10 and carotid atherosclerosis was observed. Per interquartile range increase in two-year exposure to PM2.5 (Hazard ratio (HR): 1.322, 95% confidence interval (CI): 1.219–1.434) and PM10 (HR:1.213, 95% CI: 1.116–1.319) showed the strongest association with carotid atherosclerosis, respectively. Individuals in higher quartiles of pollutants were at higher risk for carotid atherosclerosis compared with those in the lowest quartile group. Concentration response functions documented the nearly linear and nonlinear relationship and interpreted the upward trends of the risk for carotid atherosclerosis with increasing level of pollutant concentrations. Moreover, effect estimates for the mixture of pollutants and carotid atherosclerosis were larger than any of the individual pollutants (HR (95% CI) was 1.510 (1.338–1.704) and 1.613 (1.428–1.822) per quartile increase for one-year and two-year time-weighted average exposure, respectively).

Conclusions

Individual time-weighted average exposure to PM2.5 and PM10 was associated with carotid atherosclerosis. Co-exposure to ambient air pollution was also positively associated with carotid atherosclerosis.

Peer Review reports

Introduction

Cardiovascular diseases (CVDs), the leading cause of death globally, lead to approximately 32% of global deaths (https://www.who.int/en/news-room/fact-sheets/detail/cardiovascular-diseases(cvds)). Atherosclerosis is the primary underlying pathology of CVDs [1]. The development of atherosclerosis is the result of total exposure to individual and environmental risk factors interacting with genetic susceptibility [2]. Ambient pollution has been identified as the most important environmental risk factor [3]. Ambient particulate matter with aerodynamic diameter < 10 μm (PM10) and < 2.5 μm (PM2.5) are considered to be the main adverse factor of human health [4] because of the broad range of toxic substances they contain. Exposure to ambient particulate matter (PM) would contribute to the progression of subclinical disease and the development of atherosclerosis [5, 6].

Carotid atherosclerosis is a reliable and well-developed clinical indicator of CVDs [7]. The progression would trigger 20% of the whole risk of ischemic stroke [8] and 50% deaths of atherosclerotic diseases [9]. Carotid intima-media thickness (CIMT) or carotid plaque are well-established quantitative measurements for carotid atherosclerosis and better biomarkers for predicting the risk of CVDs and atherosclerotic vascular disease [10]. Association between PM and atherosclerotic lesion has been explored by animal models and population-based studies [11, 12], but the results were not generally consistent. Long-term exposure to PM was demonstrated as an important risk factor for CIMT in both cross-sectional and longitudinal studies among late midlife women [13] or women transitioning through menopause [14], adolescents or young adults [15], children [16], general population [17, 18] and even HIV-positive adolescents [19]. But such association was not found in some other studies for either PM2.5 [20, 21] or PM10 [15]. PM2.5 and PM10 exposure was associated with the prevalence and progression of carotid plaque, but not with CIMT [22]. Besides, investigators reported that the significant association of PM2.5 and PM10 with CIMT or carotid plaque disappeared after adjusting for cardiovascular and socioeconomic risk factors [23].

Estimation of individual exposure to PM2.5 and PM10 simply based on monitoring stations would induce exposure misclassification [13, 21]. The significant association between PM2.5 and atherosclerosis disappeared after adjusting for metropolitan area [24], which interprets the implication of spatial characteristics for pollutant distribution. Land-use regression (LUR) model [25, 26] is a reliable method to estimate individual PM exposure considering both influencing factors and spatiotemporal characteristics of pollutant transmission. However, estimation from simple prediction model failed to account for individual activity characteristics and infiltrated ambient pollution. The Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air) study was well-performed with complete protocol to estimate long-term individual exposure to PM2.5 [27]. Time spent both indoors and outdoors and infiltrated ambient PM2.5 and PM10 was accounted for [21, 27]. But exposure on commuting were not comprehensively considered, which was identified as a source of personal exposure [28, 29]. It is reliable for children and the elderly, because it is unlikely for them to change the residence frequently in a certain period of time, and they spent less time outdoors compared with adults [30, 31]. However, time spent on commuting or around work place should not be ignored for adults, and exposure on days off work was different from that on workdays. Previous research had investigated daily space–time-activity-weighted model to estimate short-term individual exposure [32], and its association with cardiopulmonary outcomes changed with additional consideration of space–time activity information compared with crude estimation. This method is relatively easy to implement for short-term exposure. But as for long-term exposure, important personal day-to-day activity should also be considered. Previous studies had characterized personal exposure to PM2.5 [21, 27], however, individual time-weighted PM10 exposure was not fully calculated.

Ambient PM would induce the mortality and morbidity of sudden cardiac and cerebrovascular events by promoting the progression of atherosclerotic plaque and heightened thrombogenicity. This is a long-term progression with interacted and joint adverse effect of multiple factors [25]. Ambient air pollution is a mixture of particles and gases. These pollutants are highly correlated, and their combined effects on health exist. Double or multi-pollutant models were performed to obtain their independent effect [26,27,28,29,30,31, 33]. But the overall effect of co-exposure to pollutants cannot be interpreted, and high collinearity cannot be avoided. Quantile g-computation was presented to eliminate potential errors arising from collinearity by weighted method and obtain the total effect [34, 35].

In this study, we calculated one- and two-year time-weighted average exposures to PM2.5 and PM10 based on LUR model considering personal activity pattern (commuting mode, physical activity modes, residential and work addresses, time spent indoors, outdoors and on commuting) and infiltration factors. We aimed to investigate long-term association between time-weighted average exposure to PM2.5 and PM10 and the risk of carotid atherosclerosis, and further explore the overall effect of co-exposure to ambient pollutants on carotid atherosclerosis.

Materials and methods

Participants

Study population derived from Beijing Health Management Cohort (BHMC) study. Detailed information of physical examination was recorded each year. The detailed procedure of participant enrollment was described in Fig. 1. Related data was collected from the 2013–2014 baseline survey to December 31st, 2020 (n = 12,480). To calculate individual exposure to ambient pollutants, only participants with both residential and work addresses were included (n = 3253). We further excluded those with coronary heart disease, cancer or cerebrovascular disease (n = 184). Finally, a total number of 3069 participants were enrolled.

Fig. 1
figure 1

Flow chart about the enrollment of the participants

Full size image

Data collection

Anthropometric and laboratory measurements were conducted by trained medical professionals at baseline and for each follow-up visit. This study followed the 1964 Declaration of Helsinki and the later amendments, and was approved by the Ethics Committee of Capital Medical University (NO: 2019SY088). Written informed consent was obtained from all the participants.

Participants were supposed to take off shoes and heavy clothes to take the anthropometric measurements. Body mass index (BMI) was calculated dividing weight (kg) by the square of height (m2). Blood pressure (systolic blood pressure (SBP) and diastolic blood pressure (DBP)) was measured in the right arm in the sitting position using electronic sphygmomanometer with at least 5 min’ rest and 30 min’ interval during which time individuals were not allowed to drink any beverages containing caffeine or alcohol. The equipment should be put at the height of heart and the tested arm. Three times’ measurements were performed, and the records were averaged and applied. Mean arterial pressure (MAP) was calculated by two thirds of DBP plus one third of SBP.

Information of demographic, lifestyle, disease and medication history, including age, sex, physical activity intensity (low, moderate and high intensity), education level (lower or equal to high-school degree, college or university degree, and over or equal to postgraduate degree), excessive salt intake (> 6 g/day or ≤ 6 g/day), drinking status (current alcohol drinker), smoking status (current smoker), disease and medication history of hypertension, diabetes and dyslipidemia was collected based on questionnaire. Physical activity intensity is a question item by asking directly “What about your physical activity intensity” containing three options: ① low intensity: no exercise or mild activities, such as walking, dancing or doing Tai Chi, et al.; ② moderate intensity: including jogging, bicycling or climbing, et al.; ③ high intensity: including swimming, skipping rope or racket and balling sports, et al. Smoking status was characterized as never or former, and current smoker, and drinking status as never or former, and current drinker.

Variables calculating time-weighted average concentration including residential and work addresses were also collected based on questionnaires (Additional file A1). The amount of time spent indoors around the residential and work places commonly were recorded directly (hours) within one day. Commuting modes were categorized as ① driving a car or taking a taxi; ② walking, cycling, taking electric car or motorcycle; ③ taking a bus; ④ taking the subway. And time spent on commuting for a single travel was collected directly (hours) or based on the following options which were defined as ① less than 30 min; ② 30–60 min; ③ longer than 60 min. The specific values we applied were 30, 45 and 60 min for each option, respectively. Information about physical activity frequency was collected based on the following question: How many times did you do exercises for a week? The answers included ① one to two times per week; ② three to five times per week; ③ more than five times per week. The values of two, four and six times were applied for this question, respectively, when calculating time-weighted average exposure. The length for a single time of exercise was collected as “How long about your exercise length for a single time of exercise?”. The answers included three items: ① less than 30 min; ② 30 to 60 min; ③ more than 60 min. And 30, 45 and 60 min were finally applied separately. Activity at indoor or outdoor environment was decided based on the type of exercise. For instance, indoor activity includes yoga, swimming and balling sports in gym, et al. And outdoor activity includes jogging, climbing and bicycling, et al.

After at least 12 h’ overnight fasting, blood samples were collected and analyzed. The following indicators, including fasting blood glucose (FBG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG) and uric acid (UA), were measured by enzymatic method using a chemistry analyzer (Beckman LX 20, Beckman, Brea, CA, America) at the central laboratory.

Individual time-weighted exposure to PM

Individual exposure to ambient PM was estimated based on a satellite-based LUR model, and the model performance was shown in Table A1. Additionally considering micro-environment exposure sources and exposure time, individual time-weighted average exposure was calculated.

Prediction model

LUR model combined machine learning algorithms was performed to predict daily concentrations of ambient pollutants. Explanatory variables included aerosol optical depth (AOD), meteorological variables, land-use, normalized difference vegetation index (NDVI), traffic road, population density and elevation data.

Pollutant concentrations including PM2.5, PM10 and gaseous pollutants (sulfur dioxide (SO2), nitrogen dioxide (NO2), carbon monoxide (CO) and ozone (O3)) were obtained from Beijing National Environmental Monitoring Center (http://www.bjnemc.cn/) at 35 monitoring stations from 2013 to 2020. The recorded measurements were checked to remove negative values. Valid measurements were aggregated to obtain daily mean values during the study period at each monitoring station. Original AOD data (MCD19A2) were downloaded from the National Aeronautics and Space Administration (NASA) Moderate Resolution Imaging Spectroradiometer (MODIS) (https://ladsweb.modaps.eosdis.nasa.gov/) onboard the Terra and Aqua satellites. AOD values at 550 nm were applied. Based on the new algorithm of multi-angle implementation of atmospheric correction (MAIAC), AODs with similar accuracy for dark and vegetated areas but higher accuracy for brighter areas [36] were obtained. The product could enlarge the number of valid daily AOD values [37]. AODs with the highest quality were recommended. MOD04 and MYD04 products were applied to fill the missing values. Four main meteorological variables (daily mean temperature, two-minute-average wind speed, barometric pressure and relative humidity at 1,000 hPa) were collected from China Meteorological Administration (http://data.cma.cn). Universal kriging method with an appropriate 1-km resolution was applied to estimate missing values. Annual land-use data at a spatial resolution of 1-km in 2015, 2018 and 2020 were obtained from Resource and Environment Science and Data Center (https://www.resdc.cn/). Road traffic data were obtained through OpenStreetMap (http://download.geofabrik.de/). The total cover of each land use type (including water, forest and urban cover) and total road length in buffers within the given radius were calculated, respectively. The radiuses of buffer zones for land-use type were set as 1 km, 2 km, 3 km, 4 km and 5 km, and for road length were 100 m, 200 m, 300 m, 400 m, 500 m, 600 m, 700 m, 800 m, 900 m, 1 km, 2 km and 3 km. The 16-day NDVI data with 1 km × 1 km grid resolution (MOD13A2) was obtained from NASA official website (https://ladsweb.modaps.eosdis.nasa.gov/). ASTER GDEM Version 3 elevation data (a spatial resolution of approximate 30 m’ horizontal posting at the equator) was downloaded from NASA website (https://lpdaac.usgs.gov/products/astgtmv003/). Population density data was collected via Beijing Statistical Yearbook (http://nj.tjj.beijing.gov.cn/nj/).

Residential and work addresses were geocoded based on Maplocation (https://maplocation.sjfkai.com/) and transformed into standard coordinate system. Explanatory variables and dependent variable (pollutant concentration) were matched according to geographical position and time (day or year). Univariate regression was performed to primarily select explanatory variables linking to pollutant concentration (P value for regression coefficient < 0.05). Considering the nonlinear relationship between explanatory variables and pollutant concentration, generalized additive model (GAM) was applied to select variables with better explanatory ability. Predictors were individually added to the model, and the inclusion should not turn over the predefined direction of both individual explanatory variable and the former model. Variance inflation factor (VIF) > 4 was treated as collinearity [38]. We fitted extreme gradient boosting (XGBoost) model with tenfold cross-validation, and the prediction model yielded a mean R2 of 0.787 for PM2.5 and 0.715 for PM10 (Table A1).

Calculation of time-weighted individual concentration

We calculated one- and two-year time-weighted average exposure to ambient pollutants based on the following equations [39].

$${Y}_{air}=\left({C}_{hout}\ast{T}_{hout}+{C}_{wout}\ast{T}_{wout}+{IF}_{h}\ast\left({C}_{hout}\ast{T}_{hin}+{C}_{wout}\ast{T}_{win}\right)+{T}_{trans}\ast{IF}_{t}\ast\left({C}_{hout}+{C}_{wout}\right)/2\right)/24$$
(1)
$${Y}_{air}=\left({C}_{hout}\ast{T}_{hout}+{IF}_{h}\ast{C}_{hout}\ast{T}_{hin}+{T}_{trans}\ast{IF}_{t}\ast{C}_{hout}\right)/24$$
(2)

Specifically, Eq. (1) illustrated the method of time-weighted pollutant estimation on working days, and Eq. (2) on days off work. In the above-mentioned equations, time scale is deemed as days. In detail, \({C}_{hout}\) and \({C}_{wout}\) represent outdoor concentrations estimated based on LUR model according to individual longitudes and latitudes of geocoded residential and work addresses, respectively; \({T}_{hout}\) and \({T}_{hin}\) represent time spent outside and inside around the residences, respectively; \({T}_{wout}\) and \({T}_{win}\) represent time spent outside and inside around work places, respectively; \({T}_{trans}\) represents the time spent on commuting; \({IF}_{h}\) and \({IF}_{t}\) represent infiltration factors for ambient air pollutants infiltrating into indoors and different traffic vehicles, respectively [40,41,42]. Exposure on different traffic vehicles was estimated by the average level of infiltrated ambient pollutant around outside residential and outside work places \([\left({C}_{hout}+{C}_{wout}\right)/2]\). Missing values of time spent on commuting were filled in based on the exposure handbook among Chinese population [43] and multiple imputation method. Multivariate imputation via chained equation was performed and five imputed datasets were generated based on an imputing regression model to fill in missing values. Because we did not find suitable infiltration factor of bus for PM10, the infiltration factor of bus for PM2.5 was applied, instead. Results of correlation analysis between individual pollutant exposure estimated based on time-weighted method, residential and work address were shown in Table A2 (Additional file), and the significant association was found between PM2.5 and PM10 estimation of time-weighted method and at residential address.

For those diagnosed as carotid atherosclerosis during the follow-up, one- (365-day) and two-year (730-day) time-weighted average exposure to PM2.5 and PM10 was calculated before the date of diagnosis for carotid atherosclerosis. Similarly, three-, four- and five-year average exposure were also obtained. Individuals without carotid atherosclerotic lesion during the follow-up were treated as censoring, and time-weighted concentration was calculated before the date of the last follow-up.

Definition of carotid arteriosclerosis

Carotid artery ultrasonography measurement was performed for the diagnosis of carotid atherosclerosis by trained physicians using a 3.5-MHz transducer (logic Q700 MR, GE, Milwaukee, WI, USA) at baseline and during follow-up. Participants were required to be in a supine position. Shortly, bilateral examination on both right and left carotid arteries was carried out using a standardized approach at common carotid artery, external carotid artery and internal carotid artery. Carotid bifurcation and bulb were also measured. CIMT was determined by measuring the distance between the echo fronts of the media adventitia and lumen intima at four different locations (right, left, near walls and far walls). Carotid plaque was characterized as focal thickening of IMT greater than 1.5 mm or a protrusion greater than 50% in relation to neighboring segments [44]. Carotid atherosclerosis was defined by the presence of carotid plaque or focal thickening of the carotid artery wall [45]. Detailed procedure and criteria were rigorously performed according to Mannheim consensus [46].

Statistical analysis

Baseline characteristics were presented as number (percentage) for categorical variables and mean (standard deviation (SD)) for normal distributed continuous variables or median (interquartile range (IQR)) for continuous variables that did not follow normal distribution. Both of the two pollutants (PM2.5 and PM10) were categorized into 4 groups according to the quartiles. Stepwise Cox regression models were performed to explore the association between the risk of carotid atherosclerosis and individual exposure to PM2.5 and PM10 in both continuous and categorical form. Hazard ratios (HRs) and 95% confidence intervals (CIs) were presented for per IQR increase in pollutant concentrations. Model 1 was adjusted for age and sex. Model 2 further controlled education level, smoking status, drinking status, physical activity intensity, excessive salt intake and medication history. Model 3 was additionally adjusted for BMI, MAP, FBG, UA, LDL-C, HDL-C and TG based on model 2. Since the exposure–response trend could be evaluated by modelling PM as a linear term, spline was set for violations of linearity assumption to accommodate potential non-linear relationship. Multivariable adjusted restricted cubic spline model was conducted to assess the exposure–response relationship between time-weighted average exposure to PM2.5 and PM10 and the risk of carotid atherosclerosis. The number of knots between 3 and 7 were selected and finally decided by comparing Akaike information criterion. Figure results were derived from restricted cubic spline model with 3 degrees of freedom.

Quantile g-computation was performed to evaluate the total effect of multiple ambient pollutant exposures on carotid atherosclerosis. Quantile g-computation is a parametric, generalized-linear model-based method for estimating mixing effect that makes use of g computation. It also supports non-linear and non-additive effects of individual and mixture pollutants, as well as exposures that are associated with study outcome in different directions or with positive or negative weights. Coefficients > 0 for the individual pollutant indicates positive weight; coefficients < 0 indicates negative weight. The overall estimation is the sum of coefficients for pollutants corresponding to the change in risk of carotid atherosclerosis. In this work, quantile g-computation is used to describe the overall risk of carotid atherosclerosis controlling for covariates and give separate importance of each pollutant (PM2.5, PM10, SO2, NO2, CO and O3). The estimated overall coefficients of all exposures increasing by one quartile at the same time and weights that indicate the contribution of each individual component to the overall estimate were obtained.

We performed several sensitivity analyses described as follows: ① Individuals with only work address were included in model 3 (n = 11840); ② Individuals without medication history of hypertension, diabetes and hyperlipidemia were included in model 3 (n = 2606); ③ Individual time-weighted average exposure to ambient pollutants was estimated without consideration of infiltration factors; ④ We further adjusted for other pollutant and ran two-pollutant models according to the results of correlation analysis (Significant correlation was deemed as correlation coefficient larger than 0.6 or smaller than -0.6, and P value < 0.05) (Table A3). ⑤ Individuals with complete imputed information of calculating time-weighted concentrations were included for the analysis (n = 836). We also investigated associations between PM2.5 and PM10 and risk of carotid atherosclerosis for subgroups defined by age (≥ 60 years or < 60 years), sex (male or female), LDL-C (≥ 3.4 mmol/L or < 3.4 mmol/L), TG (≥ 1.7 mmol/L or < 1.7 mmol/L), HDL-C (≥ 1.0 mmol/L or < 1.0 mmol/L), BMI (≥ 24 kg/m2 or < 24 kg/m2) and hyperuricemia (yes or no). Subgroup factors for dyslipidemia were collected based on guidelines among Chinese [47]. Hyperuricemia was defined as ≥ 360 μmol/L in women and ≥ 420 μmol/L in men.

For all analyses, a two-tailed P value < 0.05 was considered to be statistically significant. Stepwise Cox regression analysis was performed by SAS version 9.4 (SAS Institute, Cary, North Carolina, USA), and multivariable adjusted restricted cubic spline model and quantile g-computation by R version 4.1.1.

Results

A total of 3069 individuals were included, and 981 (31.96%) individuals diagnosed as carotid atherosclerosis during the study were observed (Table 1). At baseline, the median (IQR) age of the participants was 42 (34–49) years. Nearly half of the participants were men (46.27%). Most of the participants got the college or university degree (76.34%) and had a relatively low physical activity intensity (64.03%). Most participants were nonsmokers (81.13%), nondrinkers (62.63%), and had a relatively lower salt intake (72.94%). A total of 463 individuals had a medication history of hyperlipidemia (n = 117), hypertension (n = 366) or diabetes (n = 117). The median concentration of one- and two-year time-weighted average exposure to PM2.5 was 41.899 µg/m3 and 41.997 µg/m3, and 32.526 µg/m3 and 32.495 µg/m3 for PM10 exposure, respectively (Table 2). Concentration of time-weighted average exposure to PM10 is lower than PM2.5 due to the application of infiltration factors. Air pollutant exposures within the same time window were positively correlated, except for O3, which was negatively or positively correlated with PM2.5 and PM10 (Table A2).

Table 1 Characteristics of the study participants
Full size table
Table 2 Descriptive statistics of individual time-weighted exposure to ambient PM
Full size table

The association between time-weighted average exposure to PM2.5 and PM10 and the risk of carotid atherosclerosis was significant (Table A4-A5, Fig. 2). Time-weighted individual one-year average exposure to PM2.5 and PM10 was associated with 28.1% and 16.9% increased risk of carotid atherosclerosis based on the crude model. Similar results were obtained for two-year exposure (HR (95% CI) for PM2.5 and PM10 were 1.335 (1.246–1.429) and 1.219 (1.130–1.314), respectively). The relatively smaller effect sizes were obtained for extended models adjusting for important covariates. The highest quartile (versus the lowest quartile) of PM2.5 was associated with the highest risk of carotid atherosclerosis. Effect sizes [HR (95% CI)] for increasing quartiles of time-weighted one-year average exposure to PM2.5 were 1.237 (1.003–1.525), 1.473 (1.202–1.806) and 1.656 (1.352–2.027), and of two-year exposure were 1.296 (1.049–1.601), 1.536 (1.252–1.883) and 1.890 (1.541–2.317) of the three higher quartile groups, respectively. The association of PM10 with carotid atherosclerosis were also presented when modeling PM10 as categorical variable for one- and two-year exposure [HR (95% CI): 1.251 (1.013–1.545), 1.370 (1.112–1.689) and 1.272 (1.033–1.567) for one-year exposure, and 1.188 (0.959–1.472), 1.576 (1.282–1.937) and 1.467 (1.192–1.806) for two-year exposure of the three higher quartile groups, respectively]. Clearly, the association between ambient PM and carotid atherosclerosis existed with further adjustment for personal behavior characteristics and biochemical indicators.

Fig. 2
figure 2

Relationship between individual time-weighted average exposure to PM2.5 and PM10 and carotid atherosclerosis. PM2.5, particulate matter with aerodynamic diameter < 2.5 μm; PM10, particulate matter with aerodynamic diameter < 10 μm; Q, quartile; HR, hazard ratio; CI, confidence interval; Ref, reference

Full size image

Multivariable Cox regression models with restricted cubic splines showed that the relationship between time-weighted average exposure to PM2.5 and the risk of carotid atherosclerosis appeared to be approximately linear across the exposure concentration (Fig. 3). The slope of curves did not show a level off trend at higher level of concentrations in the estimated range. The positive linear relationship was observed of two-year average exposure to ambient PM10 over the range of exposure (Fig. 3). The overall upward trend of the risk of carotid atherosclerosis was also shown with increasing PM10 concentration (Fig. 3).

Fig. 3
figure 3

Exposure–response curves for the relationship between individual time-weighted average exposure to PM2.5 and PM10 and carotid atherosclerosis. A and B demonstrated the trend for the risk of carotid atherosclerosis with the increasing level of one- and two-year average exposure to PM2.5, respectively. C and D described the exposure–response relationship between one- and two-year average exposure to PM10 and the risk of carotid atherosclerosis. PM2.5, particulate matter with aerodynamic diameter < 2.5 μm; PM10, particulate matter with aerodynamic diameter < 10 μm; HR, hazard ratio

Full size image

Individual time-weighted average exposure to each pollutant contributed to either positive or negative partial effects (Fig. A1) estimated by quantile g-computation approach. The results showed the strong positive weight coefficients of SO2, PM2.5 and CO. The total positive association between co-exposure to mixed ambient pollutants (PM2.5, PM10, SO2, NO2, CO and O3) and the risk of carotid atherosclerosis were demonstrated. The risk of carotid atherosclerosis increased by 51% with a change in mixture of time-weighted one-year average exposure to ambient air pollutants by one quartile [HR (95% CI): 1.510 (1.338–1.704)]. The overall mixture effects were driven by SO2 (66.7%), followed by CO (13.8%), PM2.5 (13.4%), and O3 (6.1%). The risk of carotid atherosclerosis was also strongly associated with two-year time-weighted co-exposure to ambient air pollution [HR (95% CI): 1.613 (1.428–1.822)], and the importance of the pollutants was similar to that of one-year exposure [SO2 (65.0%), followed by CO (14.7%), PM2.5 (13.7%), and O3 (6.6%), respectively]. For PM10 a strong negative weight was obtained. Detail results of quantile g-computation are shown in Table A6.

Results of sensitivity analysis revealed that our results were robust. The significant relationship between PM2.5 and PM10 exposure and the risk of carotid atherosclerosis was additionally shown (Table A7-A8). Different exposure period did not change the significant association between PM2.5 and PM10, except for five-year average exposure. Excluding individuals with missing values of time-weighted method, the results were robust when modeling pollutant as continuous variable. The association between PM2.5 exposure and carotid atherosclerosis were independent of NO2, O3 and CO. However, the significant association between PM10 and carotid atherosclerosis disappeared additionally adjusting for CO.

We did not find significant associations between the risk of carotid atherosclerosis and time-weighted average exposure to PM2.5 and PM10 among subjects who were 60 years or older. We observed the stronger associations between PM2.5 and carotid atherosclerosis among males and participants with hyperuricemia, normal level of HDL-C and TG and abnormal level of LDL-C and BMI. Subgroups, which included females and participants with hyperuricemia, normal level of HDL-C and abnormal level of LDL-C, TG and BMI, were more sensitive to PM10 exposure with a stronger association with carotid atherosclerosis (Table A9-A10).

Discussion

We calculated one- and two-year individual time-weighted average exposure to PM2.5 and PM10 based pollutant estimation from LUR model and a time-weighted method. PM2.5 and PM10 were identified as risk factors for carotid atherosclerosis. About 32.2% and 21.3% increased risk of carotid atherosclerosis were found with two-year exposure to PM2.5 and PM10, respectively. The consistent results were obtained for one-year exposure and treating PM as categorical variables. The linear relationship between PM2.5 exposure and the risk of carotid atherosclerosis was observed. Such relationship was also shown for PM10 exposure. Robust results were demonstrated by a variety of sensitivity analyses. The overall adverse effect of time-weighted co-exposure to ambient pollutants on the risk of carotid atherosclerosis was obtained. Our results suggest that higher levels of PM2.5 and PM10 exposure would increase the risk of carotid atherosclerosis.

The detailed mechanisms of PM-induced atherosclerotic lesion were not clearly and fully illustrated, but several processes would potentially describe the progression according to mechanistic researches and experiments. Exposure to PM2.5 and PM10 would promote the release of inflammatory cytokines and reactive oxygen species, induce peroxidative stress and inflammatory process, and further cause endothelial malfunction and vascular remodeling [48, 49]. Tonne et al. had investigated the association between oxidative potential of PM10 and CIMT, but it was not statistically significant [50]. More original studies should be performed to investigate the relationship between oxidation and inflammation derived from PM and carotid atherosclerosis. We observed larger effect sizes for PM2.5 than PM10. It can be explained by greater pulmonary deposition and differences in sources and constituents. Ambient PM2.5 primarily emits from fossil combustion, and PM10 consists of higher fractions of particles, such as suspended dust. Larger particles had relatively less infiltration efficiency into indoor environment. Our study also demonstrated the significant association between multiple pollutant exposures and the risk of carotid atherosclerosis based on quantile g-computation. Evidence about the association of ambient pollutants with carotid atherosclerosis was enriched.

It was evidenced that increased atherosclerotic burden was observed in areas with higher levels of air pollutants [51]. PM2.5 estimated by AOD adjusting for visibility and relative humidity was proved to be associated with atherosclerosis [52]. The significant association of atherosclerosis with interpolated PM2.5 was demonstrated among treatment groups of four clinical trials, but not among the whole population or other subgroups [33]. Research derived from Whitehall II study reported the significant association between 52 weeks’ average exposure to PM10 estimated by a geostatistical spatiotemporal model with the percent change of CIMT before ultrasound scan [50] without adjustment for important blood biochemical indicators. Study among randomized participants of two rigorous clinical trials also found the significantly higher percent change of CIMT with PM2.5 exposure estimated by interpolation model, but some covariates, such as blood lipids, were not controlled [53]. Long-term association between PM2.5 and PM10 derived from gaussian dispersion model and the prevalence of carotid plaques was significant for an exposed time of lag 15–18 years at three-year follow-up [22]. Among healthy participants of Multicultural Community Health Assessment Trial, the association between PM2.5 predicted based on LUR model and the progression of carotid artery atherosclerosis was not significant [26]. A study of midlife population from a prospective cohort found the significant association between estimated individual PM2.5 exposure (industry and household emissions) and carotid atherosclerosis, but such relationship was not shown with PM2.5 and PM10 derived from traffic and heating [23]. Several well-performed cohorts were combined and the overall effect was pooled [54], but the significant associations between PM2.5 and PM10 and CIMT were not observed.

The results were not totally consistent. Characteristics of the study population is the main source of heterogeneity. BHMC study includes most of individuals taking health examination every year in Beijing. Participants in the same examination group hold the similar socioeconomical characteristics and activity patterns at work. We controlled important metabolic and behavior indicators to explore the independent association of PM with carotid atherosclerosis. Accurate estimation of individual exposure to pollutants would help quantify the actual relationship between PM and carotid atherosclerosis. Compared with previous studies [13, 18, 22], we applied LUR model and a time-weighted method to estimate individual exposure [21, 27]. The association between PM10 and PM2.5 and cardiopulmonary outcomes became statistically significant when space–time-activity-weighted exposure was calculated. But daily space–time-activity-weighted estimates of PM10 was not significantly associated with reduced FVC [32]. Daily activity mode was suspected to confound the relationship of pollution exposures with atherosclerosis, but physical activity as an independent variable measured by metabolic-equivalents-of-task was not a mediator [32]. Physical activity and ambient air pollution have a complex interaction on health condition, and their interacted or combined effect on health should be further assessed. Nonetheless, daily-activity-related information deserves more attention to obtain more accurate estimation of individual exposure to ambient pollution.

Our results demonstrated that women and those with higher level of TG, LDL-C, BMI, UA and HDL-C had higher risk of carotid atherosclerosis when exposing to PM10. While men and individuals with lower level of TG were elusively sensitive to PM2.5 with higher risk of carotid atherosclerosis. It was once reported that a greater susceptibility for atherosclerotic lesion of ambient PM were interpreted among individuals with dyslipidemia [13], overweight [15] and hyperuricemia. One possible explanation for this finding is the weaker association of PM with carotid atherosclerosis compared with cardiometabolic risk factors. Pre-existing health conditions would also have an effect. Additionally, such results should be explained with caution because of the small sample size. The observed vulnerability of men may be described as more time spent outdoors, larger airways and enhanced deposition of PM.

Our study has several advantages. First, individual exposure to PM was calculated based on LUR model at both residential and work addresses and a time-weighted method, which can better represent personal exposure. Ambient and indoor PM are highly correlated, but infiltrated indoor ambient PM should not be ignored. Second, quantile g-computation was applied to control collinearity of the pollutants, and the overall adverse effects of individual exposure to multiple pollutants were obtained in our study. Our study enriched evidences about the total positive effect of multiple pollutant exposures, which was less explored before.

Some limitations would affect our findings. Household PM is also an important source of individual pollution, such as fuel use and tobacco smoking, but it was not considered in our study due to limited data. Infiltration factors of bus for PM10 were not clearly investigated according to current studies. Thus, infiltration factor of bus for PM2.5 was applied to approximately represent that of PM10, which would induce bias. We collected individual activity information only at the baseline questionnaire. Additionally, small samples in subgroup analysis and enrolling participants only in Beijing would inhibit the result generalization. Due to exclusion of individuals without residential or work address to calculate time-weighted average exposure, the sample size became smaller, which would result in potential selection bias. The negative weight for PM10 is difficult to interpret, possibly reflecting co-linearity with SO2, which suggests that there is a mixture effect, and the overall effect may be more robust than the weights of individual pollutants. Further researches are required to explore the relationship between real personal exposure to PM and the prevalence and progression of carotid atherosclerosis considering follow-up information among larger sample of population in multicenter studies. The overall effect of co-exposure to ambient air pollutants on carotid atherosclerosis and their respective contribution to overall effect should also be investigated.

Conclusions

In conclusion, our study provides population-based evidences about long-term association between time-weighted average exposures to PM2.5 and PM10 and the risk of carotid atherosclerosis. Our findings highlight the role of PM as a risk factor for carotid atherosclerosis proved by the linear or nonlinear exposure–response curves. Besides, the overall adverse effect of multiple pollutants is also proved. These findings contribute to both policy-making on improvement of ambient pollution and prevention of atherosclerosis-related diseases.

Availability of data and materials

The datasets are not publicly available because of the private information, but could be accessed from the corresponding author with reasonable request.

Abbreviations

PM:

Particulate matter

PM10 :

Particulate matter with aerodynamic diameter < 10 μm

PM2.5 :

Particulate matter with aerodynamic diameter < 2.5 μm

HR:

Hazard ratio

CI:

Confidence interval

CVDs:

Cardiovascular diseases

CIMT:

Carotid intima-media thickness

LUR:

Land-use regression

MESA Air:

Multi-Ethnic Study of Atherosclerosis and Air Pollution

BHMC:

Beijing Health Management Cohort

BMI:

Body mass index

MAP:

Mean arterial pressure

FBG:

Fasting blood glucose

HDL-C:

High-density lipoprotein cholesterol

LDL-C:

Low-density lipoprotein cholesterol

TG:

Triglycerides

UA:

Uric acid

SO2 :

Sulfur dioxide

NO2 :

Nitrogen dioxide

CO:

Carbon monoxide

O3 :

Ozone

SD:

Standard deviation

IQR:

Interquartile range

References

  1. Lu H, Daugherty A. Recent highlights of ATVB atherosclerosis. Arterioscler Thromb Vasc Biol. 2015;35(3):485–91. https://doi.org/10.1161/ATVBAHA.115.305380.

    Article  CAS  Google Scholar 

  2. Lusis AJ. Atherosclerosis Nature. 2000;407(6801):233–41. https://doi.org/10.1038/35025203.

    Article  CAS  Google Scholar 

  3. GBD 2015 Risk Factors Collaborators. Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016;388(10053):1659–1724. https://doi.org/10.1016/S0140-6736(16)31679-8. Erratum in: Lancet. 2017;389(10064):e1.

  4. WHO Regional Office for Europe. Review of evidence on health aspects of air pollution – REVIHAAP project: technical report. Copenhagen: WHO Regional Office for Europe; 2013.

    Google Scholar 

  5. Stachyra K, Kiepura A, Olszanecki R. Air pollution and atherosclerosis - a brief review of mechanistic links between atherogenesis and biological actions of inorganic part of particulate matter. Folia Med Cracov. 2017;57(3):37–46.

    Google Scholar 

  6. Newby DE, Mannucci PM, Tell GS, Baccarelli AA, Brook RD, Donaldson K, et al. Expert position paper on air pollution and cardiovascular disease. Eur Heart J. 2015;36(2):83–93b. https://doi.org/10.1093/eurheartj/ehu458.

    Article  CAS  Google Scholar 

  7. Kim H, Kim JY, Min PK, Yoon YW, Lee BK, Hong BK, et al. Outcomes and associated factors of discrepant coronary and carotid atherosclerosis. Int Heart J. 2020;61(6):1142–9. https://doi.org/10.1536/ihj.20-318.

    Article  CAS  Google Scholar 

  8. Finn C, Giambrone AE, Gialdini G, Delgado D, Baradaran H, Kamel H, et al. The association between carotid artery atherosclerosis and silent brain infarction: a systematic review and meta-analysis. J Stroke Cerebrovasc Dis. 2017;26(7):1594–601. https://doi.org/10.1016/j.jstrokecerebrovasdis.2017.02.028.

    Article  Google Scholar 

  9. Roger VL, Go AS, Lloyd-Jones DM, Adams RJ, Berry JD, Brown TM, et al. Heart disease and stroke statistics–2011 update: a report from the American Heart Association. Circulation. 2011;123(4):e18–209. https://doi.org/10.1161/CIR.0b013e3182009701. Epub 2010 Dec 15. Erratum in: Circulation. 2011 Feb 15;123(6):e240. Erratum in: Circulation. 2011 Oct 18;124(16):e426.

    Article  Google Scholar 

  10. Naqvi TZ, Lee MS. Carotid intima-media thickness and plaque in cardiovascular risk assessment. JACC Cardiovasc Imaging. 2014;7(10):1025–38. https://doi.org/10.1016/j.jcmg.2013.11.014.

    Article  Google Scholar 

  11. Soares SR, Carvalho-Oliveira R, Ramos-Sanchez E, Catanozi S, da Silva LF, Mauad T, et al. Air pollution and antibodies against modified lipoproteins are associated with atherosclerosis and vascular remodeling in hyperlipemic mice. Atherosclerosis. 2009;207(2):368–73. https://doi.org/10.1016/j.atherosclerosis.2009.04.041.

    Article  CAS  Google Scholar 

  12. Künzli N, Perez L, von Klot S, Baldassarre D, Bauer M, Basagana X, et al. Investigating air pollution and atherosclerosis in humans: concepts and outlook. Prog Cardiovasc Dis. 2011;53(5):334–43. https://doi.org/10.1016/j.pcad.2010.12.006.

    Article  Google Scholar 

  13. Duan C, Talbott E, Brooks M, Park SK, Broadwin R, Matthews K, et al. Five-year exposure to PM2.5 and ozone and subclinical atherosclerosis in late midlife women: The Study of Women’s Health Across the Nation. Int J Hyg Environ Health. 2019;222(2):168–76. https://doi.org/10.1016/j.ijheh.2018.09.001.

    Article  CAS  Google Scholar 

  14. Duan C, Talbott EO, Broadwin R, Brooks M, Matthews K, Barinas-Mitchell E. Residential exposure to PM2.5 and Ozone and progression of subclinical atherosclerosis among women transitioning through menopause: the study of Women’s Health Across the Nation. J Womens Health (Larchmt). 2019;28(6):802–11. https://doi.org/10.1089/jwh.2018.7182.

    Article  Google Scholar 

  15. Chen SY, Hwang JS, Chan CC, Wu CF, Wu C, Su TC. Urban air pollution and subclinical atherosclerosis in adolescents and young adults. J Adolesc Health. 2022;71(2):233–8. https://doi.org/10.1016/j.jadohealth.2022.03.004.

    Article  Google Scholar 

  16. Guo YL, Ampon RD, Hanigan IC, Knibbs LD, Geromboux C, Su TC, et al. Relationship between life-time exposure to ambient fine particulate matter and carotid artery intima-media thickness in Australian children aged 11–12 years. Environ Pollut. 2021;291:118072. https://doi.org/10.1016/j.envpol.2021.118072.

    Article  CAS  Google Scholar 

  17. Su TC, Hwang JJ, Shen YC, Chan CC. Carotid intima-media thickness and long-term exposure to traffic-related air pollution in middle-aged residents of Taiwan: a cross-sectional study. Environ Health Perspect. 2015;123(8):773–8. https://doi.org/10.1289/ehp.1408553.

    Article  CAS  Google Scholar 

  18. Aguilera I, Dratva J, Caviezel S, Burdet L, de Groot E, Ducret-Stich RE, et al. Particulate matter and subclinical atherosclerosis: associations between different particle sizes and sources with carotid intima-media thickness in the SAPALDIA Study. Environ Health Perspect. 2016;124(11):1700–6. https://doi.org/10.1289/EHP161.

    Article  Google Scholar 

  19. Toe S, Nagy M, Albar Z, Yu J, Sattar A, Nazzinda R, et al. Ambient air pollution is associated with vascular disease in Ugandan HIV-positive adolescents. AIDS. 2022;36(6):863–70. https://doi.org/10.1097/QAD.0000000000003186.

    Article  CAS  Google Scholar 

  20. Farzan SF, Habre R, Danza P, Lurmann F, Gauderman WJ, Avol E, et al. Childhood traffic-related air pollution and adverse changes in subclinical atherosclerosis measures from childhood to adulthood. Environ Health. 2021;20(1):44. https://doi.org/10.1186/s12940-021-00726-x.

    Article  CAS  Google Scholar 

  21. Kaufman JD, Adar SD, Barr RG, Budoff M, Burke GL, Curl CL, et al. Association between air pollution and coronary artery calcification within six metropolitan areas in the USA (the Multi-Ethnic Study of Atherosclerosis and Air Pollution): a longitudinal cohort study. Lancet. 2016;388(10045):696–704. https://doi.org/10.1016/S0140-6736(16)00378-0. Epub 2016; 24. Erratum in: Lancet. 2016 Aug 13;388(10045):660.

    Article  CAS  Google Scholar 

  22. Sommar JN, Norberg M, Grönlund C, Segersson D, Näslund U, Forsberg B. Long-term exposure to particulate air pollution and presence and progression of carotid artery plaques - A northern Sweden VIPVIZA cohort study. Environ Res. 2022;211:113061. https://doi.org/10.1016/j.envres.2022.113061.

    Article  CAS  Google Scholar 

  23. Hasslöf H, Molnár P, Andersson EM, Spanne M, Gustafsson S, Stroh E, et al. Long-term exposure to air pollution and atherosclerosis in the carotid arteries in the Malmö diet and cancer cohort. Environ Res. 2020;191:110095. https://doi.org/10.1016/j.envres.2020.110095.

    Article  CAS  Google Scholar 

  24. Kim SY, Sheppard L, Kaufman JD, Bergen S, Szpiro AA, Larson TV, et al. Individual-level concentrations of fine particulate matter chemical components and subclinical atherosclerosis: a cross-sectional analysis based on 2 advanced exposure prediction models in the multi-ethnic study of atherosclerosis. Am J Epidemiol. 2014;180(7):718–28. https://doi.org/10.1093/aje/kwu186.

    Article  Google Scholar 

  25. Peralta AA, Schwartz J, Gold DR, Vonk JM, Vermeulen R, Gehring U. Quantile regression to examine the association of air pollution with subclinical atherosclerosis in an adolescent population. Environ Int. 2022;164:107285. https://doi.org/10.1016/j.envint.2022.107285.

    Article  CAS  Google Scholar 

  26. Gan WQ, Allen RW, Brauer M, Davies HW, Mancini GB, Lear SA. Long-term exposure to traffic-related air pollution and progression of carotid artery atherosclerosis: a prospective cohort study. BMJ Open. 2014;4(4):e004743. https://doi.org/10.1136/bmjopen-2013-004743.

    Article  Google Scholar 

  27. Kaufman JD, Adar SD, Allen RW, Barr RG, Budoff MJ, Burke GL, et al. Prospective study of particulate air pollution exposures, subclinical atherosclerosis, and clinical cardiovascular disease: The Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air). Am J Epidemiol. 2012;176(9):825–37. https://doi.org/10.1093/aje/kws169.

    Article  Google Scholar 

  28. Wang M, Sampson PD, Sheppard LE, Stein JH, Vedal S, Kaufman JD. Long-term exposure to ambient ozone and progression of subclinical arterial disease: the multi-ethnic study of atherosclerosis and air pollution. Environ Health Perspect. 2019;127(5):57001. https://doi.org/10.1289/EHP3325.

    Article  Google Scholar 

  29. Sun M, Kaufman JD, Kim SY, Larson TV, Gould TR, Polak JF, et al. Particulate matter components and subclinical atherosclerosis: common approaches to estimating exposure in a Multi-Ethnic Study of Atherosclerosis cross-sectional study. Environ Health. 2013;3(12):39. https://doi.org/10.1186/1476-069X-12-39.

    Article  Google Scholar 

  30. Iannuzzi A, Verga MC, Renis M, Schiavo A, Salvatore V, Santoriello C, et al. Air pollution and carotid arterial stiffness in children. Cardiol Young. 2010;20(2):186–90. https://doi.org/10.1017/S1047951109992010.

    Article  Google Scholar 

  31. Wilker EH, Mittleman MA, Coull BA, Gryparis A, Bots ML, Schwartz J, et al. Long-term exposure to black carbon and carotid intima-media thickness: the normative aging study. Environ Health Perspect. 2013;121(9):1061–7. https://doi.org/10.1289/ehp.1104845.

    Article  CAS  Google Scholar 

  32. Cole-Hunter T, de Nazelle A, Donaire-Gonzalez D, Kubesch N, Carrasco-Turigas G, Matt F, et al. Estimated effects of air pollution and space-time-activity on cardiopulmonary outcomes in healthy adults: a repeated measures study. Environ Int. 2018;111:247–59. https://doi.org/10.1016/j.envint.2017.11.024.

    Article  CAS  Google Scholar 

  33. Künzli N, Jerrett M, Garcia-Esteban R, Basagaña X, Beckermann B, Gilliland F, et al. Ambient air pollution and the progression of atherosclerosis in adults. PLoS One. 2010;5(2):e9096. https://doi.org/10.1371/journal.pone.0009096.  Erratum in: PLoS One. 2010;5(3). https://doi.org/10.1371/annotation/21f6b02b-e533-46ca-9356-86a0eef8434e.

    Article  CAS  Google Scholar 

  34. Li R, Li S, Pan M, Chen H, Liu X, Chen G, et al. Physical activity counteracted associations of exposure to mixture of air pollutants with mitochondrial DNA copy number among rural Chinese adults. Chemosphere. 2021;272:129907. https://doi.org/10.1016/j.chemosphere.2021.129907.

    Article  CAS  Google Scholar 

  35. Li R, Hou J, Tu R, Liu X, Zuo T, Dong X, et al. Associations of mixture of air pollutants with estimated 10-year atherosclerotic cardiovascular disease risk modified by socio-economic status: The Henan Rural Cohort Study. Sci Total Environ. 2021;793:148542. https://doi.org/10.1016/j.scitotenv.2021.148542.

    Article  CAS  Google Scholar 

  36. Lyapustin A, Wang Y, Laszlo I, Kahn R, Korkin S, Remer L, Levy R, Reid J. Multiangle implementation of atmospheric correction (MAIAC): 2. Aerosol algorithm. J Geophys Res. 2011;116:D03211. https://doi.org/10.1029/2010JD014986.

    Article  Google Scholar 

  37. He Q, Gao K, Zhang L, Song Y, Zhang M. Satellite-derived 1-km estimates and long-term trends of PM2.5 concentrations in China from 2000 to 2018. Environ Int. 2021;156:106726. https://doi.org/10.1016/j.envint.2021.106726. Epub 2021 Jun 25. PMID: 34175778.

    Article  CAS  Google Scholar 

  38. Amini H, Taghavi-Shahri SM, Henderson SB, Naddafi K, Nabizadeh R, Yunesian M. Land use regression models to estimate the annual and seasonal spatial variability of sulfur dioxide and particulate matter in Tehran. Iran Sci Total Environ. 2014;1(488–489):343–53. https://doi.org/10.1016/j.scitotenv.2014.04.106. Epub 2014 May 16 PMID: 24836390.

    Article  CAS  Google Scholar 

  39. Liu X, Dong M, Wang J, Chen D, Xiao J, Zeng W, Li X, Hu J, He G, Ma W, Liu T. A weekly time-weighted method of outdoor and indoor individual exposure to particulate air pollution. MethodsX. 2019;18(6):2439–42. https://doi.org/10.1016/j.mex.2019.10.013.PMID:31720233;PMCID:PMC6838529.

    Article  Google Scholar 

  40. Tsai DH, Wu YH, Chan CC. Comparisons of commuter’s exposure to particulate matters while using different transportation modes. Sci Total Environ. 2008;405(1–3):71–7. https://doi.org/10.1016/j.scitotenv.2008.06.016. Epub 2008 Jul 26 PMID: 18657850.

    Article  CAS  Google Scholar 

  41. Yan C, Zheng M, Yang Q, Zhang Q, Qiu X, Zhang Y, Fu H, Li X, Zhu T, Zhu Y. Commuter exposure to particulate matter and particle-bound PAHs in three transportation modes in Beijing. China Environ Pollut. 2015;204:199–206. https://doi.org/10.1016/j.envpol.2015.05.001. Epub 2015 May 14 PMID: 25978352.

    Article  CAS  Google Scholar 

  42. Hu Y, Yao M, Liu Y, Zhao B. Personal exposure to ambient PM2.5, PM10, O3, NO2, and SO2 for different populations in 31 Chinese provinces. Environ Int. 2020;144:106018. https://doi.org/10.1016/j.envint.2020.106018. Epub 2020 Aug 6. PMID: 32771828.

    Article  CAS  Google Scholar 

  43. Duan X, Zhao X, Wang B, Chen Y, Cao S. Highlights of the Chinese exposure factors handbook. Elsevier; 2015. https://www.researchgate.net/profile/Xiaoli-Duan/publication/294645842_Highlights_of_the_Chinese_Exposure_Factors_HandbookAdults/links/56c2c22008aeedba0567fac1/Highlights-of-the-Chinese-Exposure-Factors-HandbookAdults.pdf.

  44. Lou Y, Li X, Cao L, Qin P, Shi J, Zhang Y, Wang C, et al. LDL-cholesterol to HDL-cholesterol ratio discordance with lipid parameters and carotid intima-media thickness: a cohort study in China. Lipids Health Dis. 2020;19(1):141. https://doi.org/10.1186/s12944-020-01324-5.

    Article  CAS  Google Scholar 

  45. Bos S, Duvekot MH, Ten Kate GR, Verhoeven AJ, Mulder MT, Schinkel AF, et al. Carotid artery plaques and intima medial thickness in familial hypercholesteraemic patients on long-term statin therapy: A case control study. Atherosclerosis. 2017;256:62–6. https://doi.org/10.1016/j.atherosclerosis.2016.12.005.

    Article  CAS  Google Scholar 

  46. Touboul PJ, Hennerici MG, Meairs S, Adams H, Amarenco P, Bornstein N, et al. Mannheim carotid intima-media thickness and plaque consensus (2004-2006-2011). An update on behalf of the advisory board of the 3rd, 4th and 5th watching the risk symposia, at the 13th, 15th and 20th European Stroke Conferences, Mannheim, Germany, 2004, Brussels, Belgium, 2006, and Hamburg, Germany, 2011. Cerebrovasc Dis. 2012;34(4):290–6. https://doi.org/10.1159/000343145.

    Article  Google Scholar 

  47. Joint committee issued Chinese guideline for the management of dyslipidemia in adults. [2016 Chinese guideline for the management of dyslipidemia in adults]. Zhonghua Xin Xue Guan Bing Za Zhi. 2016; 24;44(10):833–853.  https://doi.org/10.3760/cma.j.issn.0253-3758.2016.10.005. Chinese.

  48. Miller MR, Shaw CA, Langrish JP. From particles to patients: oxidative stress and the cardiovascular effects of air pollution. Future Cardiol. 2012;8(4):577–602. https://doi.org/10.2217/fca.12.43.

    Article  CAS  Google Scholar 

  49. Quan C, Sun Q, Lippmann M, Chen LC. Comparative effects of inhaled diesel exhaust and ambient fine particles on inflammation, atherosclerosis, and vascular dysfunction. Inhal Toxicol. 2010;22(9):738–53. https://doi.org/10.3109/08958371003728057.

    Article  CAS  Google Scholar 

  50. Tonne C, Yanosky JD, Beevers S, Wilkinson P, Kelly FJ. PM mass concentration and PM oxidative potential in relation to carotid intima-media thickness. Epidemiology. 2012;23(3):486–94. https://doi.org/10.1097/EDE.0b013e31824e613e.

    Article  Google Scholar 

  51. Oikonomou E, Lazaros G, Mystakidi VC, Papaioannou N, Theofilis P, Vogiatzi G, et al. The association of air pollutants exposure with subclinical inflammation and carotid atherosclerosis. Int J Cardiol. 2021;1(342):108–14. https://doi.org/10.1016/j.ijcard.2021.07.056.

    Article  Google Scholar 

  52. Woo KS, Chook P, Hu YJ, Lao XQ, Lin CQ, Lee P, et al. The impact of particulate matter air pollution (PM2.5) on atherosclerosis in modernizing China: a report from the CATHAY study. Int J Epidemiol. 2021;50(2):578–88. https://doi.org/10.1093/ije/dyaa235.

    Article  CAS  Google Scholar 

  53. Künzli N, Jerrett M, Mack WJ, Beckerman B, LaBree L, Gilliland F, et al. Ambient air pollution and atherosclerosis in Los Angeles. Environ Health Perspect. 2005;113(2):201–6. https://doi.org/10.1289/ehp.7523.

    Article  CAS  Google Scholar 

  54. Perez L, Wolf K, Hennig F, Penell J, Basagaña X, Foraster M, et al. Air pollution and atherosclerosis: a cross-sectional analysis of four European cohort studies in the ESCAPE study. Environ Health Perspect. 2015;123(6):597–605. https://doi.org/10.1289/ehp.1307711.

    Article  Google Scholar 

Download references

Acknowledgements

The authors thanked for all of the participants in BHMC study.

Funding

This work was supported by the National Natural Science Foundation of China (grant numbers: 81872708, 82073668).

Author information

Authors and Affiliations

  1. School of Public Health, Capital Medical University, No.10 Xitoutiao, You’anmenWai, Fengtai District, Beijing, 100069, China

    Ze Han, Xiaoyu Zhao, Zongkai Xu, Jinqi Wang, Rui Jin, Yueruijing Liu, Zhiyuan Wu, Jie Zhang, Xiuhua Guo & Lixin Tao

  2. Beijing Municipal Key Laboratory of Clinical Epidemiology, Beijing, 100069, China

    Ze Han, Xiaoyu Zhao, Zongkai Xu, Jinqi Wang, Rui Jin, Yueruijing Liu, Jie Zhang, Xiuhua Guo & Lixin Tao

  3. Department of Public Health, School of Medical and Health Sciences, Edith Cowan University, Perth, Australia

    Zhiyuan Wu

  4. Department of Mathematics and Statistics, La Trobe University, Melbourne, 3086, Australia

    Xia Li

Authors
  1. Ze Han
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Xiaoyu Zhao
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Zongkai Xu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jinqi Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Rui Jin
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Yueruijing Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Zhiyuan Wu
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Jie Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Xia Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Xiuhua Guo
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Lixin Tao
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Z.H.: Conceptualization, Methodology, Software, Formal analysis, Investigation, Data Curation, Writing-Original Draft; X.Z.: Formal analysis, Writing-Original Draft; Z.X.: Software, Formal analysis; J.W.: Validation; R.J.: Validation; Y.L.: Writing-Original Draft; J.Z.: Supervision; Z.W.: Writing-Review & Editing; X.L.: Writing-Review & Editing; X.G.: Validation, Writing-Review & Editing; L.T.: Conceptualization, Methodology, Investigation, Writing-Review & Editing, Data Curation, Supervision, Project administration, Funding acquisition. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Lixin Tao.

Ethics declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committees of Capital Medical University (approval number: 2019SY088). And written informed consent was obtained from all of the participants.

Consent for publication

Not applicable.

Competing interests

The authors declare they have nothing to disclose.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1:

Additional file A1 illustrated information for several parts of questionnaire for BHMC study. Table A1. Information about predictive model accuracy for each pollutant. PM2.5, particulate matter with aerodynamic diameter <2.5 μm; PM10, particulate matter with aerodynamic diameter <10 μm; NO2, nitrogen dioxide; SO2, sulfur dioxide; O3, ozone; CO, carbon monoxide; RMSE, root mean square error; MAE, mean absolute error. Table A2. Correlation analysis between individual exposure to PM2.5 and PM10 estimated based on time-weighed method, residential address and work address * P values <0.05 for Spearman correlation coefficients. PM2.5, particulate matter with aerodynamic diameter <2.5 μm; PM10, particulate matter with aerodynamic diameter <10 μm. Table A3. Spearman’s correlation coefficients between individual time-weighted average exposure to ambient pollutants * P values <0.05 for correlation coefficients. PM2.5, particulate matter with aerodynamic diameter <2.5 μm; PM10, particulate matter with aerodynamic diameter <10 μm; NO2, nitrogen dioxide; SO2, sulfur dioxide; O3, ozone; CO, carbon monoxide. Table A4. Estimated risk for carotid atherosclerosis associated with individual time-weighted average exposure to PM2.5 a Model 1: adjusted for age and gender.  b Model 2: adjusted for variables in model 1, as well as education level, smoking status, drinking status, physical activity intensity, excessive salt intake and medication history of hypertension, diabetes and hyperlipidemia.  c Model 3: adjusted for variables in model 2 plus BMI, MAP, UA, FBG, LDL-C, HDL-C and TG. Ref, Reference; PM2.5, particulate matter with aerodynamic diameter <2.5 μm; Q, quartile; HR, hazard ratio; CI, confidence interval; IQR, interquartile range; FBG, fasting blood glucose; MAP, mean arterial pressure; BMI, body mass index; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; UA, uric acid. Table A5. Estimated risk for carotid atherosclerosis associated with individual time-weighted average exposure to PM10 a Model 1: adjusted for age and gender. b Model 2: adjusted for variables in model 1, as well as education level, smoking status, drinking status, physical activity intensity, excessive salt intake and medication history of hypertension, diabetes and hyperlipidemia. c Model 3: adjusted for variables in model 2 plus BMI, MAP, UA, FBG, LDL-C, HDL-C and TG. Ref, Reference; PM10, particulate matter with aerodynamic diameter <10 μm; Q, quartile; HR, hazard ratio; CI, confidence interval; IQR, interquartile range; FBG, fasting blood glucose; MAP, mean arterial pressure; BMI, body mass index; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; UA, uric acid. Table A6. Estimated risk for carotid atherosclerosis in multi-pollutant models according to quantile g-computation. Covariates in model 3 were adjusted for. Coefficient indicates the separate effect among the total effect for individual pollutants. Weight indicates the importance of each pollutant which has the same effect direction. Estimate indicates the combined estimated coefficient. PM2.5, particulate matter with aerodynamic diameter <2.5 μm; PM10, particulate matter with aerodynamic diameter <10 μm; NO2, nitrogen dioxide; SO2, sulfur dioxide; O3, ozone; CO, carbon monoxide; HR, hazard ratio; CI, confidence interval. Table A7. Sensitivity analyses for the association between PM2.5 exposure and carotid atherosclerosis Variables was adjusted for in model 3. - Because of high correlation and collinearity, two-pollutant model was not conducted. a The estimation of individual exposure did not consider pollutant infiltration from outdoor to indoor environment. b Individuals with complete work address (n=11843) were included in the analysis. c Individuals with missing values of time-weighted variables were excluded (n=836). Ref, Reference; PM2.5, particulate matter with aerodynamic diameter <2.5 μm; Q, quartile; HR, hazard ratio; CI, confidence interval; NO2, nitrogen dioxide; SO2, sulfur dioxide; O3, ozone; CO, carbon monoxide. Table A8. Sensitivity analyses for the association between PM10 exposure and carotid atherosclerosis Variables was adjusted for in model 3. - Because of high correlation and collinearity, two-pollutant model was not conducted. a The estimation of individual exposure did not consider pollutant infiltration from outdoor to indoor environment. b Individuals with complete work address (n=11843) were included in the analysis. c Individuals with missing values of time-weighted variables were excluded (n=836). Ref, Reference; PM10, particulate matter with aerodynamic diameter <10 μm; Q, quartile; HR, hazard ratio; CI, confidence interval; NO2, nitrogen dioxide; O3, ozone; CO, carbon monoxide. Table A9. Subgroup analyses exploring the association between individual PM2.5 exposure and carotid atherosclerosis Variables was adjusted for in model 3. Ref, Reference; PM2.5, particulate matter with aerodynamic diameter <2.5 μm; Q, quartile; HR, hazard ratio; CI, confidence interval; BMI, body mass index; LDL-C, low density lipoprotein cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol. Table A10. Subgroup analyses exploring the association between individual PM10 exposure and carotid atherosclerosis Variables was adjusted for in model 3. Ref, Reference; PM10, particulate matter with aerodynamic diameter <10 μm; Q, quartile; HR, hazard ratio; CI, confidence interval; BMI, body mass index; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol. Fig. A1. Combined effect of personal time-weighted average exposure to mixed ambient pollutants on the risk of carotid atherosclerosis. A and B represented the weight of one- and two-year co-exposure to ambient air pollutants on the risk of carotid atherosclerosis, respectively, according to quantile g-computation regression. PM10, particulate matter with aerodynamic diameter <10 μm; PM2.5, particulate matter with aerodynamic diameter <2.5 μm; NO2, nitrogen dioxide; SO2, sulfur dioxide; O3, ozone; CO, carbon monoxide.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Han, Z., Zhao, X., Xu, Z. et al. Associations of time-weighted individual exposure to ambient particulate matter with carotid atherosclerosis in Beijing, China. Environ Health 22, 45 (2023). https://doi.org/10.1186/s12940-023-00995-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12940-023-00995-8

Keywords

Environmental Health

ISSN: 1476-069X