Association Between Long-Term Air Pollution and Increased Blood Pressure and Hypertension in China

As the major risk factor for cardiovascular disease and other vascular diseases,^1–3 high blood pressure (BP) and hypertension have been identified as the most important causes of disability and the leading risk factors for death in the world, causing an estimated 7.5 million deaths per year (13% of all deaths).⁴ Because chronically elevated BP is an important determinant for the development and progression of atherosclerosis, the general trend in most developing countries of the increasing prevalence of hypertension is of public health concern.⁵ It has been suggested that this increase cannot be explained by genetic factors and changes in lifestyle and diet structure alone but that environmental factors, including ambient air pollution, may play a role as well.⁶

In recent years, several studies have investigated the short-term effects of ambient air pollutants in the development of high BP and hypertension among humans, but the results of these studies are inconstant. Evidence from studies conducted in the United States,^7–10 Brazil,¹¹ and Taiwan¹² indicated that short-term increases in particulate matter (PM), organic carbon, black carbon, and ozone (O₃) air pollution lead to acute but transient increase in arterial BP and hypertension prevalence. However, results from 1 recent study conducted on 9238 nonsmoking adults >30 years of age indicated that short-term exposure to particles with an aerodynamic diameter ≤10 µm (PM₁₀), sulfur dioxide (SO₂), nitrogen dioxides (NO₂), CO, and O₃ consistently reduced the systolic BP (SBP) and pulse pressure, whereas the diastolic BP (DBP) was increased by SO₂, NO₂, and O₃.¹³ Stronger associations existed among men, participants >60 years of age, those with hypertension, and those living in an industrial township. Also, Mordukhovich et al¹⁴ observed positive associations not between BP and PM, but between BP and black carbon. Another recent controlled human inhalation study reported that exposure to diesel exhaust was associated with a rapid and measureable increase in SBP but not DBP.¹⁵ Furthermore, some authors even reported that there was a small significant decrease in DBP and SBP and hypertension in association with particulate air pollution.¹⁶ The reasons for inconsistent effects of ambient air pollution on BP in different parts of the world are unclear but may result from spatial and temporal variability in pollution sources and composition.^17,18

Compared with short-term effects of air pollution, there is little information on the relation between chronic exposure to air pollution and prevalent hypertension. Inconsistent results have also been reported on the association between incident hypertension and air pollution. For instance, Coogan et al¹⁹ showed in a 10-year follow-up for incident hypertension and diabetes mellitus of black women from Los Angeles that NO_X (but less PM_2.5) borderline increased the risk of becoming hypertensive. However, Sørensen et al²⁰ reported that long-term exposure (1 and 5 years) to NO_X was not associated with incident self-reported hypertension in a Danish cohort. We hypothesized that at the population level, long-term exposure to ambient air pollution is associated with higher BP and higher prevalence of hypertension among humans. In this study, we tested this hypothesis in the 11 Districts Chinese Study, a cross-sectional study of air pollution and adult health in a large, well-characterized population-based sample, residing in northeast China, where there are wide differences in inter- and intracity gradients and ambient pollutant levels, offering a valuable opportunity to assess the associations between exposure and response.

There were a total of 28 830 individuals randomly selected from 33 communities of the 3 cities, of whom 24 845 completed the survey and examination to give an overall response rate of 86.2%. Among the 24 845 participants analyzed, the average age was 45.59 years (SD =13.31 years) and 12 661 (50.1%) were men. For all participants, the prevalence rate of hypertension was 34.84%. The characteristics of the survey participants in this study, stratified by sex, are shown in Table 1. Men and women were found to be statistically different across all demographical and behavioral variables except age and race.

Table 2 presents the arithmetic mean for all 4 pollutants PM₁₀, NO₂, SO₂, and O₃ measured between 2006 and 2008, and compares them with the World Health Organization (WHO) guidelines and Chinese National Ambient Air Quality Standards. The table shows that PM₁₀ and SO₂ levels exceed WHO guidelines in all districts, whereas 90.9% of all districts exceed PM₁₀ levels and 27.3% of all districts exceed SO₂ levels dictated by the Chinese National Ambient Air Quality Standards; 18.2% of all districts exceed WHO guidelines for NO₂. Additionally, 18.2% of all 11 Chinese districts exceed NO₂ levels dictated by both WHO guidelines and those of the Chinese National Ambient Air Quality Standards. O₃ levels in all districts were below the guidelines set by the WHO and those of the Chinese National Ambient Air Quality Standards.

A GLIMMIX model was used to assess the association between air pollution and hypertension. The results are displayed in Table 3. For all participants, we observed significant associations between prevalence rate of hypertension and PM₁₀ (odds ratio [OR], 1.12; 95% confidence interval [CI], 1.08–1.16), SO₂ (OR, 1.11; 95% CI, 1.04–1.18), and O₃ (OR, 1.13; 95% CI, 1.06–1.20). However, when stratified by sex, these relations were obtained only in men, and none of the pollutants was found to be associated with prevalence of hypertension among women. For instance, in the single pollutant model analysis, positive associations were identified for hypertension with interquartile range (IQR) of PM₁₀ (OR, 1.17; 95% CI, 1.06–1.30), SO₂ (OR, 1.19; 95% CI, 1.05–1.34), and O₃ (OR, 1.21; 95% CI, 1.04–1.38) only in men, but not in women. We found no association of NO₂ with prevalence of hypertension in either men or women.

We also found significant associations of long-term exposure of ambient air pollutants with arterial BP (Table 4). For all subjects, all the ambient air pollutants except for NO₂ were significantly associated with SBP and DBP levels. When stratified by sex, these significant associations were limited to men. For example, in the single pollutant model, DBP and SBP increased by 0.44 mm Hg (95% CI, 0.10–0.78 mm Hg) and 0.98 mm Hg (95% CI, 0.45–1.52 mm Hg) per 19 μg/m³ increase in 3-year mean PM₁₀, 0.50 mm Hg (95% CI, 0.20–0.79) and 1.13 mm Hg (95% CI, 0.67–1.60 mm Hg) per 20 μg/m³ increase in 3-year mean SO₂, 0.58 mm Hg (95% CI, 0.24–0.91 mm Hg) and 1.05 mm Hg (95% CI, 0.52–1.58 mm Hg) per 22 μg/m³ increase in 3-year mean O₃ in men but not in women, respectively. The estimated ORs varied by age group (Table 5). The ORs among participants ≥65 years of age seemed larger than those in the other 2 age groups with an exception of NO₂. The insignificant effects in the ≥65 years of age group might be attributable to the smaller sample size (2234) compared with the sample size (18 698) in the <55 years of age group.

The findings confirmed our hypothesis of the presence of an adverse association between long-term ambient air pollution and hypertension, and more generally a positive association with arterial SBP and DBP. These effects were seen with certain pollutants (PM₁₀, SO₂, and O₃) and observed only in men, suggesting that sex may be an effect modifier.

In contrast to an abundance of data demonstrating associations between short-term exposure to ambient air pollution and arterial BP, there is little published information on the relation between long-term exposure to ambient pollutants and measured BP or frank hypertension. We are aware of only 3 relevant human studies examining the effect of long-term air pollutants exposure on prevalence of hypertension and arterial BP.^12,26,27 The results of these studies are inconsistent. In the Heinz Nixdorf Recall Study, a population-based prospective cohort in Germany, Fuks et al²⁷ investigated the cross-sectional association of residential long-term PM_2.5 and PM₁₀ exposure with arterial BP and hypertension on 4291 participants aged 45 to 75 years, taking short-term variations of PM and long-term road traffic noise exposure into account. The results indicated that an IQR increase in PM₁₀ (3.9 μg/m³) was associated with estimated increase in mean SBP and DBP of 1.1 mm Hg (95% CI, 0.2–2.0 mm Hg) and 0.8 mm Hg (95% CI, 0.3–1.2 mm Hg), respectively. However, no association of PM with prevalence of hypertension was found.²⁷ Using the Taiwanese Survey on Prevalence of Hyperglycemia, Hyperlipidemia, and Hypertension study data, Chuang et al¹² conducted secondary analyses to investigate changes in BP associated with changes in exposure to ambient air pollutant by applying generalized additive models among 7578 participants aged 16 to 90 years. The results of that study showed no significant relation between air pollution and BP except for O₃, which revealed a 0.37 mm Hg (95% CI, 0.04–0.69 mm Hg) elevation for an IQR increase of 12.15 ppb. However, analysis of data gathered in Taiwan since 1989 among 1023 participants in Social Environment and Biomarkers of Aging Study demonstrated significant increase in BP for an IQR increase in 1-year average air pollution, where per IQR change in PM₁₀ of 48 μg/m³, SBP increased by 16.34 mm Hg (95% CI, 12.27–20.42 mm Hg) and DBP increased by 14.87 mm Hg (95% CI, 12.73–17.02 mm Hg).²⁶ Compared with these investigations, air pollutants in the current study included not only PM but also other air pollutants, such as SO₂, NO₂, and O₃, and furthermore, our study observed significant relation among PM₁₀, SO₂, and O₃ and the prevalence of hypertension. The findings in our study may be explained by the differences in spatial and temporal variability of air pollutant sources and components between the different regions, difference in the age, ethnicity, and lifestyle of the study populations, differences in exposure assessment, and so on.

Interpretation of our findings requires caution. First, we only observed marginally significant associations between NO₂ and BP and prevalence of hypertension, which is different from those reported from recent studies. Sørensen et al²⁰ reported that long-term exposure of NO_X was associated with a lower prevalence of baseline self-reported hypertension. The authors used nitrogen oxides (NOx) as ([NO] + [NO2]) because NOx is a more constant and stable measure of exposure. The differences in the reported results might be attributable to the different markers of air pollution exposure. Compared with NOx, NO₂ is not stable, which may lead to significant exposure uncertainty. In our study, ORs for PM₁₀, SO₂, and O₃ in men were very comparable. NO₂ was the only exception. The correlation analyses show that PM₁₀, SO₂, and O₃ are significantly, highly correlated with each other, whereas the correlations between NO₂ and PM₁₀, SO₂, and O₃ are much weaker. The stronger correlations among PM₁₀, SO₂, and O₃ would lead to much more similar impacts on the outcomes than NO₂. In addition, there remains some doubt whether O₃ increases SBP and prevalent hypertension and even more whether O₃ increases DBP. O₃ concentrations are highly correlated with temperature. Hoffmann et al²⁸ observed that O₃ and higher temperature were associated with decreases in SBP in a study of the short-term effects of ambient air pollution and temperature on arterial BP in 70 people with type 2 diabetes mellitus. Although the combined effects of these 2 factors are not clear, the effect of O₃ exposure on SBP in the study might be masked by an independent temperature effect on SBP. Weinhold²⁹ also speculated that the opposing effects of air pollutants and heat exposures could plausibly exist in independent ways and might not cancel one another out. Furthermore, the current study is a cross-sectional study of associations between air pollution and prevalence of BP and hypertension. We could not address temporal relation between them. Sørensen et al²⁰ assessed the associations between incidence of hypertension and long-term air pollution. The authors did not observe the associations between long-term exposure to air pollution and incident hypertension. The different study designs (prevalence versus incidence) might be the reason for the difference in observed results, in addition to the differences in sociodemographic variables, lifestyle indicators, healthy history, health care access, prescription drug use, tobacco smoke, and season.

The mechanisms by which the constituents of ambient air pollution may elevate or modify BP are not yet understood. Possible causal pathways include the elicitation of oxidative stress, systemic proinflammatory responses and the activation of pulmonary reflexes in turn leading to arterial remodeling.^26,30–33 Acute effects of PM on BP may occur via activation of pulmonary reflexes that decrease parasympathetic tone while increasing sympathetic activity in the autonomic nervous system, whereas later-onset effects of PM are probably mediated via inflammation-induced vascular reactivity.³⁰ Other investigators have also reported a link between inflammatory markers (reflected by percent neutrophils and interleukin-6) and elevated BP after long-term air pollution exposure.²⁶ Elevations in fasting glucose, ApoB, and HbA1C were associated with progression of atherosclerosis and mortality attributable to cardiovascular disease.

Other investigators have also reported a link between long-term air pollution exposure and elevations in fasting glucose, ApoB, and HbA1C—all associated with progression of atherosclerosis and mortality attributable to cardiovascular disease—and immune system activation marked by increased levels of percent circulating neutrophils and interleukin-6, thought to contribute to acute coronary syndromes.^{6, 26} In smaller vessels in the arterial tree, raised intravascular pressure increases wall stress, stimulating the activation of medial smooth muscle in arterioles and subsequent hypertrophic remodeling of resistance vessels potentiating a rise in BP as a result of increased total peripheral resistance.³¹

There is growing epidemiological evidence of the differing associations between air pollution and health for men and women, but the literature is far from consistent. Our study showed that the effect of ambient air pollution exposure on BP and prevalence rates of hypertension was evident only in men. These findings are not entirely unexpected as men may have greater pulmonary exposure to ambient air pollutants because they engage in more frequent or intense outdoor activity than women.^34–36

Finally, some limitations of the current study should be noted. First, the findings cannot establish a cause-and-effect relation among long-term ambient air pollution, BP, and hypertension because of the nature of a cross-sectional study design. We do not have any baseline readings available, nor could any temporality in the association be ascertained. Selection and information bias are possible. However, we have attempted to reduce these biases as much as possible through strict quality control and assurance procedures. Second, possible misclassification of the behavioral risk factors (smoking, drinking, and exercising) could exist because of the nature of the survey questions (yes or no). Third, we develop a variable of 3-year average concentrations to serve as surrogates for the long-term exposure to PM₁₀, SO₂, NO₂, and O₃. As stated in the method section, the measurements are more likely to reflect the background air pollution levels in a city. Traffic-related air pollution exposure data are not available to this study. This approach may lead to significant uncertainty of long-term exposure. Detailed spatial and temporal information is needed to develop a better long-term exposure. Addressing spatial variations of traffic-related air pollution effects would be one of the future directions. Fourth, there is still some doubt on whether the described associations between air pollutants and SBP or hypertension are causal. Exposure to PM_2.5 and NOx does not infrequently go together with noise, and exposure to noise (especially traffic noise) has also been incriminated as a causal factor for incident hypertension. Therefore, it would be ideal to adjust for noise when air pollution effects on SBP or hypertension are assessed. Unfortunately, no noise data are available to this study. Nevertheless, these factors are not likely to vary significantly across the 7 studied cities because these are big cities in the northeast China, where city landscapes, traffic density and patterns, and time activity patterns are quite similar. Finally, we have found a strong intercorrelation between the 3 pollutants PM₁₀, SO₂, and O₃ and a moderate correlation among NO₂, PM₁₀, and O₃. Therefore, it would be inappropriate to fit a model including all 4 pollutants. We are not able to assess multiple pollutants’ effect simultaneously in this study.

In this current study, we found an association of long-term exposure to PM, SO₂, and O₃ with increased arterial BP and established hypertension in a population-based sample. This finding supports the belief that long-term ambient air pollution exposure may promote atherosclerosis and other arterial damage with air-pollution–induced increases in BP being one possible biological pathway. However, the mechanisms by which the constituents of ambient air pollution may elevate or modify BP needs to be investigated in prospective studies.

We are grateful to those participants in this study who had been very generous with their time and assistance. We would also thank the anonymous reviewers for their very helpful comments.