Article Text

Download PDFPDF

Long-term air pollution exposure and risk factors for cardiovascular diseases among the elderly in Taiwan
  1. Kai-Jen Chuang1,2,
  2. Yuan-Horng Yan2,3,
  3. Shu-Yi Chiu2,
  4. Tsun-Jen Cheng2
  1. 1Department of Cosmetic Application and Management, St. Mary's Medicine Nursing and Management College, Sanxing Township, Yilan County, Taiwan
  2. 2Institute of Occupational Medicine and Industrial Hygiene, College of Public Health, National Taiwan University, Taipei City, Taiwan
  3. 3Department of Medical Research, Chia-Yi Christian Hospital, Chiayi City, Taiwan
  1. Correspondence to Professor Tsun-Jen Cheng, National Taiwan University, Institute of Occupational Medicine and Industrial Hygiene, National Taiwan University College of Public Health, No. 17 Xu-Zhou Road, Zhongzheng District, Taipei 10055, Taiwan; tcheng{at}ntu.edu.tw

Abstract

Objectives To investigate changes in blood pressure, blood lipids, blood sugar and haematological markers of inflammation associated with changes in long-term exposure to ambient air pollutants.

Methods We conducted secondary analyses of data on blood pressure and blood biochemistry markers from the Social Environment and Biomarkers of Aging Study in Taiwan and air pollution data from the Taiwan Environmental Protection Administration in 2000. Associations of 1-year averaged criteria air pollutants (particulate matter with aerodynamic diameters <10 μm (PM10) and <2.5 μm (PM2.5), ozone (O3), nitrogen dioxide (NO2), sulfur dioxide and carbon monoxide) with systolic blood pressure, diastolic blood pressure, total cholesterol, triglycerides, high-density lipoprotein cholesterol, fasting glucose, haemoglobin A1c (HbA1c), interleukin 6 (IL-6) and neutrophils were explored by applying generalised additive models.

Results After controlling for potential confounders, we observed that increased 1-year averaged particulate air pollutants (PM10 and PM2.5) and NO2 were associated with elevated blood pressure, total cholesterol, fasting glucose, HbA1c, IL-6 and neutrophils. Associations of increased 1-year averaged O3 with elevated blood pressure, total cholesterol, fasting glucose, HbA1c and neutrophils were also observed. In particular, our two-pollutant models showed that PM2.5 was more significantly associated with end-point variables than two gaseous pollutants, O3 and NO2.

Conclusions Changes in blood pressure, blood lipids, blood sugar and haematological markers of inflammation are associated with long-term exposure to ambient air pollutants. This might provide a link between air pollution and atherosclerotic cardiovascular diseases.

  • Long-term air pollution
  • blood pressure
  • blood lipids
  • blood sugar
  • inflammation
  • epidemiology
  • public health
  • cardiovascular
  • air pollution
  • PM10-PM2.5-ultrafine

Statistics from Altmetric.com

What this paper adds

  • The effects of long-term air pollution exposure on cardiovascular morbidity and mortality are well documented, but the underlying mechanisms remain unclear.

  • Our results add to the evidence of the associations of 1-year averaged particulate matter, ozone and nitrogen dioxides with blood pressure, total cholesterol, fasting glucose, haemoglobin A1c, interleukin 6 and neutrophils.

  • Long-term air pollution may induce inflammation, hypertension, hyperlipidaemia and hyperglycaemia, which all are critical for the progression of atherosclerotic cardiovascular diseases.

Introduction

Evidence from epidemiological studies has consistently demonstrated that long-term exposure to ambient air pollutants is associated with increased cardiovascular morbidity and mortality.1 The American Cancer Society (ACS) cohort study reported that people living in more polluted areas with high levels of particulate matter (PM) with aerodynamic diameter <2.5 μm (PM2.5) were more likely to die from cardiopulmonary diseases and lung cancer than those in less polluted areas.2 3 An association between yearly average of particulate matter with aerodynamic diameter <10 μm (PM10) and increased risk for hospitalisation for congestive heart failure or subsequent myocardial infarction has also been reported.4 To investigate the possible mechanisms linking long-term air pollution exposure with cardiovascular diseases, the ACS cohort study further reported associations of PM2.5 with mortality from specific cardiopulmonary diseases and hypothesised that the general pathophysiological pathways of PM effects may include pulmonary and systemic inflammation, acceleration of atherosclerosis and alteration of cardiac autonomic function.5 Nevertheless, evidence linking long-term air pollution exposure to major cardiovascular risks/biomarkers in humans is still limited.

A recent epidemiological study of 2978 adults from the Third National Health and Nutrition Examination Survey (NHANESIII) demonstrated a link between long-term PM10 exposure and white blood cell count (WBC), a non-specific haematological marker of inflammation.6 A recent study by Hoffmann et al also showed that long-term exposure to PM was associated with C-reactive protein (CRP) and fibrinogen. In the adjusted analysis, a cross-sectional exposure difference of 3.91 μg/m3 in PM2.5 (interdecile range) was associated in men with a 23.9% increase in CRP and a 3.9% increase in fibrinogen.7 These findings supported the hypothesis that one biological mechanism, inflammation, links air pollution and cardiovascular diseases as stated by the American Heart Association, and partially supported the suggestion that inflammation plays a role in the general pathophysiological pathways linking long-term air pollution with cardiovascular diseases.5 8 However, a recent study reported that there was no association between inflammatory markers of fibrinogen or CRP and long-term exposure to outdoor air pollution, including PM10, nitrogen dioxide (NO2), sulfur dioxide (SO2) and ozone (O3).9 Another study investigating the relationship between PM2.5 and CRP showed a slightly increased risk, but it was not significant.10 Therefore, further evidence is needed to explore the effects of long-term air pollution exposure on more specific biomarkers in humans to confirm previous epidemiological studies. To address these scientific gaps, we conducted secondary data analyses to investigate the changes in blood pressure, blood lipids, blood sugar and inflammation markers in elderly subjects associated with changes in exposure to long-term ambient air pollution.

Methods

Health data

Health data were obtained from the Social Environment and Biomarkers of Aging Study (SEBAS) in Taiwan. The SEBAS sample was based on the Survey of Health and Living Status of the Elderly in Taiwan, which has been conducted on a national sample since 1989.11 In the year 2000, one quarter of the subjects in this survey were further randomly selected for blood pressure measurement and venous blood sample collection. Data on sociodemographic characteristics including sex, age, body mass index (BMI), current smoker (yes vs no), drinking (yes vs no), living region (county/city) and physician-diagnosed diseases (hypertension, hyperlipidaemia and diabetes; yes vs no) were collected by nurses who administered a questionnaire during a home visit. This information formed the SEBAS database, which included 1023 subjects aged 54 and older.

Blood pressure and biochemistry

A detailed description of the sampling procedures and data validation on the SEBAS is given in a previous study.12 In brief, systolic and diastolic blood pressure were calculated as the average of two seated blood pressure readings (taken about 1 min apart using a mercury sphygmomanometer) during examination in hospital. Blood samples were collected and sent to Union Clinical Laboratories (UCL) in Taipei for measurement of biomarkers, including fasting glucose, glycosylated haemoglobin (HbA1c), total cholesterol, triglycerides, high-density lipoprotein cholesterol (HDL-C), interleukin 6 (IL-6) and neutrophils. In addition to the routine standardisation and calibration tests performed by the laboratory, duplicate samples for a subset of 10% of the specimens were submitted to UCL and Quest Diagnostics in the USA for analysis. Data from these duplicate analyses indicated good interlaboratory and intralaboratory reliability, with intraclass correlations of 0.80 or higher for duplicates sent to UCL and interlaboratory correlations of 0.76 or higher between results from UCL and Quest Diagnostics.

Air pollution data

Seventy-two monitoring stations operated by the Taiwan Environmental Protection Administration (EPA) throughout Taiwan measured air pollution and weather data daily. Daily concentrations of PM10, O3, NO2, SO2, carbon monoxide (CO) and temperature were used to represent 1023 residents' air pollution exposure. The average concentration of air pollutants from all monitoring stations in each county or city was calculated and assigned to each individual living in the area. All daily environmental data were matched with the sampling date of blood pressure measurement and blood collection. The environmental data averaged over 365 days before the blood pressure and blood sampling date were used to estimate yearly pollution effects on blood pressure and blood markers. A map of the study area, distribution of the participants and the location of monitoring stations are presented in figure 1.

Figure 1

Map of the study area, distribution of the participants, and location of the monitoring stations. The dimensions of the circles are proportional to the population in each county/city in Taiwan. For a more detailed description, please visit the website of the Taiwan Air Quality Monitoring Network (TAQMN) operated by the Taiwan Environmental Protection Administration (http://taqm.epa.gov.tw/taqm/en/default.aspx).

Statistical analyses

We applied linear additive models to examine the associations between air pollutants, blood pressure and blood biochemistry markers.13 The exposure variables were 1-year average PM10, PM2.5, O3, NO2, SO2 and CO, and the outcome variables were systolic blood pressure, diastolic blood pressure, fasting glucose, HbA1c, total cholesterol, triglycerides, HDL-C, IL-6 and neutrophils. Each regression model included age, sex, BMI, current smoker (yes vs no) and drinking (yes vs no). We conducted residual analysis to evaluate the sensitivity of the results to residences with significant results by estimating the models with the significant residences excluded. The models also adjusted for smooth function terms as fit by penalised cubic regression spline to reflect possible non-linear effects of continuous covariates, including visit date and yearly averaged temperature with 6 degrees of freedom to adjust for time varying influences on blood pressure and blood markers.14 All statistical analyses were performed using R Statistical Software, V.2.4.1.15 Estimates of the effects of air pollutants were scaled to interquartile ranges (IQRs), the differences between the 25th and 75th percentiles, for the appropriate day's mean air pollutants.

Results

The ages of the 1023 subjects varied widely (from 54 to 90 years); 76% of subjects were current non-smokers and 34% of the subjects had hypertension (table 1). Median resting blood pressure was 139/82 mm Hg, and median fasting glucose, HbA1c, total cholesterol, triglycerides, HDL-C, IL-6 and neutrophils were 106.8 mg/dl, 5.8%, 200.7 mg/dl, 120.9 mg/dl, 48.9 mg/dl, 1.2 pg/ml and 56.8%, respectively (table 2). Air pollution and temperature measurements are shown in table 3. The concentration ranges of the pollutants varied widely during the study period. The environmental temperatures ranged from 11.7°C to 30.4°C.

Table 1

Basic characteristics of study participants (N=1023)

Table 2

Summary statistics of biomarkers

Table 3

Summary statistics of annual air pollution and temperature levels during the study period

Associations between air pollution and blood pressure

Table 4 shows increases in blood pressure for IQR increases in the 1-year average air pollution, as estimated using linear additive modelling. We observed associations of systolic and diastolic blood pressure with increases in the 1-year average of PM10, PM2.5, O3 and NO2 after adjustment for age, sex, BMI, current smoking, drinking and smoothed functions for visit date and yearly temperature. We further considered two-pollutant models for these four air pollutants. To avoid collinearity due to high correlation, PM10 and PM2.5 were fitted separately to our two-pollutant models. The results showed that PM2.5 retained the strongest association with systolic and diastolic blood pressure among the four air pollutants after controlling for other pollutants in two-pollutant models. For an IQR increase in PM2.5, we found 32.4 mm Hg (95% CI 22.4 to 42.5) and 29.3 mm Hg (95% CI 19.2 to 39.3) increases in systolic and diastolic blood pressure, respectively, after controlling for O3. For an IQR increase in PM2.5, we found 31.1 mm Hg (95% CI 21.1 to 41.2) and 30.0 mm Hg (95% CI 18.0 to 41.9) increases in systolic and diastolic blood pressure, respectively, after controlling for NO2.

Table 4

Changes (95% CI) in blood pressure and blood markers for interquartile range increases in annual air pollution estimated by generalised additive models

Associations between air pollution and blood biochemistry markers

The pollution effects on blood lipids, blood sugar and inflammatory markers estimated by single-pollutant models are listed in table 4. The blood lipid marker total cholesterol as well as the blood sugar markers of fasting glucose and HbA1c were significantly associated with 1-year average increases in particulate air pollution, O3 and NO2. The inflammatory marker of neutrophils was also significantly associated with increases in 1-year average particulate air pollution, O3 and NO2. IL-6 was marginally (p<0.10) associated with 1-year average PM10 and NO2. By contrast, triglycerides and HDL-C were not associated with any of these air pollutants. We further considered two-pollutant models for particulate air pollution, O3 and NO2 and found that only PM2.5 retained the strongest association with blood lipid, blood sugar and inflammatory markers among four air pollutants after controlling for other pollutants in two-pollutant models. For an IQR increase in PM2.5, we found 74.1 mg/l (95% CI 55.5 to 92.7), 34.6 mg/dl (95% CI 16.5 to 52.7), 2.1% (95% CI 1.5 to 2.7) and 16.1% (95% CI 12.1 to 20.0) increases in total cholesterol, fasting glucose, HbA1c and neutrophils, respectively.

Predictably, current smoking was consistently associated with increases in both blood pressure and blood biochemistry markers. Seasonal increases in temperature were associated with increases in systolic and diastolic blood pressure. No significant association of blood pressure and blood biochemistry markers was observed with SO2 and CO. We conducted residual analysis of the linear additive models by labelling residuals and partial residuals by residence address and plotting them over PM2.5 levels. The sensitivity of the results was then evaluated by excluding the 149 residences which provided highly significant observations regarding blood pressure and blood markers. We found that changes in blood pressure and blood markers by PM2.5 were very similar to the original models and the association between these end-point variables and PM2.5 was still robust. We also reanalysed our data by log10 transforming blood pressure and blood markers to improve normality and stabilise the variance. The findings from these new models showed robust associations between PM2.5 and these end-point variables.

Discussion

We found that long-term exposure for 1 year to particulate and gaseous air pollutants was associated with inflammation, high blood pressure, and increased blood lipids and blood sugar, which all indicate an increased risk of atherosclerotic cardiovascular disease.16–18 This study provides intriguing epidemiological evidence of a link between increased levels of inflammatory markers, as reflected by neutrophils and IL-6, and elevated blood pressure and long-term air pollution exposure, and supports the hypothesised mechanisms of air pollution effects on atherosclerotic cardiovascular disease through inflammatory responses or alterations in vascular tone.8 Neutrophils are the most abundant type of WBC and are an indicator of inflammation. They have been reported to be associated with increases in long-term cardiovascular risks in the Framingham Heart Study19 and the Atherosclerosis Risk in Communities Study and are a stable marker of vascular inflammation predicting long-term cardiovascular events.20–22 Moreover, both short-term and long-term exposures to ambient PM10 have been reported to be associated with increases in WBC count.6 23 24 IL-6 is involved in regulation of the synthesis of CRP, a sensitive indicator of infection, injury and inflammation in the liver. Increased concentrations of IL-6 are associated with an increased risk of cardiovascular events and mortality.25 26 Previous studies reported that serum IL-6 is increased in healthy male subjects after exposure to increased air pollution due to forest fires. Increases in blood pressure reflect a systemic increase in vascular tone, with its adverse implications for atherosclerotic cardiovascular disease.27 It has been reported that exposure to ambient PM10 was associated with elevated systolic and diastolic blood pressure among subjects with chronic obstructive pulmonary disease.28 A chamber study of healthy adults demonstrated increased diastolic blood pressure in response to controlled exposure to PM2.5 and O3.29 One recent epidemiological study also showed higher systolic blood pressure and pulse pressure were associated with ambient levels of PM2.5.30 Taken together, these studies highlight the importance of inflammation and altered vascular tone in the pathophysiological pathways underlying the adverse cardiovascular effects of both short-term and long-term air pollution exposure.

Increased blood lipids have been identified as a major cardiovascular risk factor with a strong independent impact on atherosclerotic cardiovascular disease by the Framingham Heart Study.16 The importance of total cholesterol in atherosclerosis progression and cardiovascular death has been recognised.31 In a recent panel study in Chapel Hill, North Carolina, a increase in blood lipids among adults with asthma was reported to be induced by PM10.32 Previous animal studies also demonstrated that exposure to PM10 could increase the total amount of lipids in aortic lesions in hyperlipidaemic rabbits and the lipid content in the aortic arch in mice fed a high-fat chow diet.33 34 Moreover, long-term exposure (32 weeks) to NO2 near ambient levels was reported to be associated with increased triglycerides in obese rats.35 Such findings partially support the associations we found between long-term air pollution exposure and increased blood lipids and the hypothesised involvement of blood lipids in air pollution-mediated chronic cardiovascular effects.

Another interesting finding in our study is that long-term exposure to ambient air pollution was associated with fasting glucose and HbA1c. Fasting glucose is primarily used to determine plasma glucose level after an overnight fast, while HbA1c is used to identify average plasma glucose level over prolonged periods of time. Increased fasting glucose and HbA1c have both been recognised as major cardiovascular risk factors in atherosclerosis progression and cardiovascular death. HbA1c is also associated with oxidative stress and dysfunctional endothelium and plays an important role in the pathogenesis of atherosclerosis.18 36–38 The present study showed an association between long-term air pollution exposure and increased fasting glucose and HbA1c in an elderly population. On the basis of these findings, we conclude that there are close relationships between high blood sugar and elevated atherosclerotic risk related to air pollution exposure.

In general, our findings suggested that long-term exposure to air pollution consisting of both particulate air pollutants and gaseous pollutants was associated with increased blood pressure and blood markers. The three air pollutants responsible for these end-point variables in our single-pollutant models, that is particles, O3 and NO2, were generally consistent with the air pollutants reported in previous studies on cardiovascular effects.6 23 24 28 29 In particular, our two-pollutant models showed that PM2.5 was more significantly associated with the end-point variables than two gaseous pollutants, O3 and NO2. Our results indicated that PM2.5 related to traffic emissions was more significantly associated with cardiovascular effects than other gaseous air pollutants in Taiwan.

We recognise that our secondary data analysis has limitations. Because of the lack of small-scale spatial information, we were unable to relate the spatial difference in health data to other related features of air pollution exposure beyond the varying air pollutants levels across areas. Exposure misclassification or measurement error may bias study outcomes toward either positive or null results.39 Moreover, the size of the presented effects is relatively high, showing an increase in blood pressure of more than 30 mm Hg per interquartile increase in PM2.5.29 30 Similarly, very high effects are also found for total cholesterol, fasting glucose and HbA1c. We have adjusted for individual-level confounders including age, gender, BMI, current smoking and drinking status. However, we could not rule out the possibility of unmeasured confounding due to various physical and psychosocial stressors and residential environment factors associated with both cardiovascular risks and ambient air pollution levels such as socio-economic status, complete smoking status, cumulative smoking exposure, noise exposure, occupational exposures, physical activity, blood pressure medication and lipid lowering therapy, because data on these variables are not available in SEBAS.1 5 8 Regardless of these limitations, our secondary data analyses still generally support the hypothesis that long-term air pollution exposure is associated with increases in neutrophils, IL-6, blood pressure, total cholesterol, fasting glucose and HbA1c, all atherosclerotic risk factors in the elderly.

Conclusion

The consistent associations of long-term exposure to PM10, PM2.5, NO2 and O3 with cardiovascular effects in the elderly imply the presence of long-term air pollution-induced inflammation, hypertension, hyperlipidaemia and hyperglycaemia, which are all involved in the progression of atherosclerotic cardiovascular diseases.

References

Footnotes

  • This study is based on data from the 2000 Social Environment and Biomarkers of Aging Study, provided by the Bureau of Health Promotion, Department of Health, R.O.C (Taiwan). The descriptions or conclusions herein do not represent the viewpoint of the Bureau

  • Funding This work was supported, in part, by grants (NSC-095-SAF-I-564-602-TMS and NSC-097-EPA-M-002-001) from the Taiwan National Science Council.

  • Competing interests None.

  • Patient Consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.