Hypertension is the most prevalent chronic disease among the elderly (1) and has led to adverse cardiovascular diseases (CVDs), such as hypertensive heart disease and stroke, and even death (2). Worldwide, as the elderly population continues to grow, the disease burden attributed to hypertension and its complications are increasing (3, 4) so it is imperative to identify risk factors for hypertension and promote prevention among the elderly.

In addition to well-established associations with lifestyle and heredity, it is thought that environmental pollutants also contribute to the occurrence of hypertension (5, 6). In particular, the inhalation of air containing fine particulate matter (PM), especially sizes ≤2.5 um (PM_2.5) has been reported as a probable antecedent driver of the increase in blood pressure and incidence of hypertension (7–12). Overall, although the association between PM_2.5 exposure and hypertension has been reported in several studies, gaps in knowledge remain. For example, elderly people may be more susceptible to PM_2.5, due to higher rates of CVDs and given the decline in organ function associated with age (13, 14). However, relevant studies rarely focus on the elderly, especially those aged over 80 years (15–19).

Furthermore, inconsistent results have been reported in previous studies. For instance, based on a cross-sectional study involving 27,752 Taipei City residents (mean age 74.8 years), Chen et al. found that PM_2.5 exposure was not significant associated with diastolic blood pressure (DBP) and none of the air pollutants were associated with changes in systolic blood pressure (SBP) (16). In addition, cross-sectional study designs and small-scale sample sizes limit the generalizability of the results to the elderly population (17, 19–21).

Lastly, many studies have been conducted in regions and countries, such as Taiwan and the USA (16, 21), with good air quality defined by PM_2.5 concentrations lower than the World Health Organization recommendations (10 μg/m³). It has been suggested that the conclusions drawn from studies conducted in other regions might not be suitable for generalization to China (11, 22). China has some of the heaviest air pollution in the world with an average annual population-weighted PM_2.5 exposure of 52.7 μg/m3 in 2017 (23). More importantly, PM_2.5 has been proven to be a modifiable factor that contributes to cardiovascular morbidity and mortality (24). Thus, considering the research gaps and public health significance, it is necessary to study and verify the effects of PM_2.5 on hypertension among the elderly to provide more effective interventions for the elderly and medical resource allocation.

In the present prospective cohort study based on the national representative Chinese Longitudinal Healthy Longevity Survey (CLHLS), we aimed to examine the association between PM_2.5 exposure and the risk of hypertension incidence among Chinese elderly. Secondarily, we aimed to further explore the modifying effects of PM_2.5 exposure on hypertension incidence and identify vulnerable sub-populations.

At baseline, 7,432 participants were enrolled in the study. During the 10-year follow-up, the total number of person-years was 24,222 and the incidence of hypertension was 8.5 per 100 person-years. The demographic characteristics, lifestyles, and health status of the participants from the CLHLS at baseline are presented in Table 1. The mean age of the participants was 87.7 (±11.5) years. Of the elderly, 56.4% were female and 62.6% were uneducated. Over half of them had no pension. Of the participants, 80.8% lived in rural areas and 83.7% lived with family members. More participants came from western/central China than from eastern China. A total of 17.6 and 18.4% of them had smoking and drinking habits, respectively. A total of 27.4% of the participants reported that they had a habit of exercising at the present. Only 1.7% and 5.8% of the patients reported having diabetes and heart disease, respectively. A total of 52.7% of the participants had a BMI between 18.5 and 23.9. Most of them without function disability. In our study, the range of PM_2.5 exposure was from 3.8 to 134.8 μg/m³. The median 1-year averages were PM_2.5 = 52.7, P₂₅ = 41.1 μg/m³, and P₇₅ = 70.1 μg/m³.

The distribution of residential PM_2.5 exposure of the 706 counties (districts) units at baseline is shown by a statistical map (Figure 2). The geographical distribution of PM_2.5 concentration was higher in the north than the south, and higher inland than the coastal areas. In addition, the areas with high PM_2.5 concentrations (>61 μg/m³) included the Sichuan Basin, Beijing-Tianjin-Hebei region, Shandong province, Hunan province, and Hubei province.

The Cox proportional hazards model with 4 degrees of freedom (the minimum AIC value) showed that there was a non-linear association between PM_2.5 exposure and hypertension with 3.8 μg/m³ as the reference group since it was the lowest concentration of PM_2.5 exposure (Figure 3). We observed strong evidence of linearity at concentrations below 50.1 μg/m³ and a stable increase in hazard risk between concentrations of 50.1 and 60.1 μg/m³. A steeper slope was found at 60.1 μg/m³ concentration and higher. Specifically, the hypertension risk was minor decreased over 107.6 μg/m³, but the effect was negligible because very little sample (accounting for 0.2% of total sample) over this value.

The incidence of hypertension in the first quartile, Q2, Q3, and Q4 were 9.3, 8.8, 7.3, and 8.5 per 100 person-years, respectively (Table 2). The results of the effects of PM_2.5 exposure and hypertension incidence using a random-effects Cox proportional hazards model are presented in Table 3. Adjusting for potential confounding variables such as age, sex, educational attainment, and pension et al., we found that a 1 μg/m³ increment in PM_2.5 concentration was associated with hypertension incidence [adjusted HR (AHR): 1.01, 95% CI: 1.01–1.02] and per 10 μg/m³ increment in PM_2.5, the concentration was associated with hypertension incidence (AHR: 1.12; 95% CI: 1.09–1.16). Similar results were observed according to quartiles, where higher concentrations were strongly related with hypertension incidence (Q2 AHR: 1.31, 95% CI: 1.13–1.51; Q3 AHR: 1.35, 95% CI: 1.15–1.60; Q4 AHR: 1.83, 95% CI: 1.53–2.17). Trend analysis showed that there was a linear trend between the PM_2.5 concentration quartile and hypertension incidence (P trend <0.001).

In subgroup analyses, after adjusting for related confounding variables, the effect estimate for each 10 μg/m³ increment in PM_2.5 concentration was significantly greater in individuals with without a pension, living in a rural setting, and residing in central/western China (P-value for modification effect <0.05 and significant HR value). The detailed modification effects are shown in Table 4.

We conducted a 10-year prospective cohort study to examine the association between PM_2.5 exposure and the risk of hypertension incidence among Chinese elderly. Our results indicated that long-term exposure to PM_2.5 was significantly associated with hypertension incidence among elderly individuals aged 65–116 years. We found that per 1 and 10 μg/m³ increase in PM_2.5, the HR of hypertension incidence increased by 1 and 12% among the elderly, respectively. Similarly, there were strong positive links between Q2, Q3, and Q4 with PM_2.5 exposure and hypertension incidence. Moreover, we observed that individuals with certain characteristics were more likely to be affected by PM_2.5 exposure based on subgroup analysis.

We observed that the average exposure from 1-year before the event had the strongest effect on the incidence of hypertension, which was shorter than previous studies on the association between PM_2.5 exposure and health outcomes have reported. For instance, some studies reported that a 3-year average PM_2.5 exposure was most closely related to mortality (32, 37). Similarly, other studies have verified that an exposure period of <2 years might greatly influence cardiovascular health outcomes (38, 39). Thus, our results indicated that the impact of PM_2.5 exposure was greater than we expected. The findings of our study suggest that interventions to improve air quality may reduce the incidence of hypertension within a short period.

According to the shape of the dose-response relationship between PM_2.5 and hypertension incidence, we found that there was no safe threshold for PM_2.5 exposure (Figure 3). Some studies have detected a threshold for PM_2.5 exposure and negative health outcomes (22, 40), for example, a J-shaped association existed between PM_2.5 exposure with a threshold concentration of 33 mg/m³, and function disability among the elderly. Other studies report that there is no safe threshold for PM_2.5 exposure when considering asthma mortality and all-cause mortality risk (32, 41). Currently, the evidence for safe PM_2.5 exposure with respect to hypertension incidence is limited. According to our findings, even with exposure to low levels of PM_2.5, the risk of hypertension should not be overlooked. From a public policy perspective, reducing each one unit concentration of PM_2.5 may be effective for protecting the elderly. Therefore, stricter national policy actions are recommended to improve air quality.

The adverse effects of PM_2.5 exposure and increased risk for hypertension found in our study are in line with previous studies conducted among middle-aged and elderly people under 80 years of age (19, 21). For example, Wu et al. used a database including 20,927 middle-aged and older participants and reported that the increase in hypertension incidence risk per unit PM_2.5 exposure was about 4.8 and 6.3% higher among males and females, respectively (19). Lin et al. enrolled 59,456 adults aged 50 years and older from four cohorts in China and showed that each 10 μg/m³ increase in PM_2.5 concentration increased the HR of hypertension incidence by 14 % (17). In addition, the current study showed that the HRs for Q2, Q3, and Q4 for PM_2.5 exposure were relatively higher than those in other published studies (12). One explanation for this finding was that these participants in our study were older than other studies (mean age was 87.7 years in our paper). Expert proposes that the ability to adapt to air pollution exposure may be reduced due to aging and health status (22). However, one study that investigated elderly individuals residing in Taiwan found no significant association between PM_2.5 exposure and hypertension (16). Discrepant results were also observed in a Black Women's Health Study that found PM_2.5 exposure was not significantly associated with measured hypertension among middle-aged and elderly women (9). In the present study, the mean of PM_2.5 exposure concentration (55.3 μg/m³) was relatively higher than PM_2.5 concentration from other studies conducted in the USA (8.6 μg/m³) (7), Seoul (38.9 μg/m³) (42), Europe (range: 5.0–8.9 μg/m³) (43), and a cross-sectional study from India (33.0 μg/m³) (11). Our finding was supported by a prior paper reported that the risk of PM_2.5 exposure was declining in most western developed countries; in contrast, the risk of PM_2.5 exposure was prominent in developing countries (44). Since 2013, the Chinese government has carried out a pilot project to monitor PM_2.5 concentration in 33 major cities and implemented an air quality standard for PM_2.5 concentration in 2016 (45). In 2019, the PM_2.5 concentration in 22 of 31 major cities in mainland China exceeded the annual limitation concentration for China (35.0 μg/m³) and 31 of them exceeded the WHO recommendation (10.0 μg/m³) (46). Overall, the regional development patterns and air pollutant concentrations varies across countries and regions, which may explain some of the heterogeneity of PM_2.5 exposure; the inconsistency between our results and those of other studies may also be caused by differences in time spent outdoors, population structures, and accessibility to health care (32).

To date, relevant studies have been proposed to explain the impacts of PM_2.5 exposure on blood pressure through several mechanisms. One study posits that PM_2.5 is taken into the human body by direct translocation through the olfactory bulb, leading to inflammatory responses and oxidative stress (47). Another proposes that inhaled particles destroy the balance in the autonomic nervous system, inducing a sympathetic response, followed by arterial vasoconstriction (48). Finally, it may be that long-term exposure to PM_2.5 affect the systemic hemodynamics of the body due to endothelial injury or dysfunction, which is also considered a risk factor for hypertension (10). In summary, the beneficial efforts to reduce PM_2.5 emission on hypertension among the elderly would be important.

Our findings implied that the association between each 10-μg/m³ increment in PM_2.5 exposure and hypertension was modified by different characteristics of the elderly. We found that HR seemed more apparent in individuals without pensions. One study concluded that the health impact of PM_2.5 on individuals with lower socioeconomic status was slightly greater than that of others (49). In our study, most of the participants were uneducated and had no pension, so they were probably more likely to spend their time engaging in outdoor manual labor that more easily suffered from PM_2.5 exposure. Additionally, the subgroup analysis showed that respondents who lived in rural areas and central/western China were more likely to have hypertension. According to the distribution of PM_2.5 concentration in mainland China in this study (Figure 2), we found most of the central or western areas such as Szechwan, Hunan, and Hubei provinces, and the Beijing-Tianjin-Hebei region had higher air pollutant levels. It is important to note that these areas are traditional industrial districts with a high population density. In rural areas, the PM_2.5 concentration was higher than in urban areas probably due to residential energy materials such as coal and biomass fuels. Accordingly, several approaches, such as clean energy use in rural areas and upgrading of traditional industries in traditional industrial districts, would improve air quality and decrease the risk of hypertension among the elderly.

Some limitations of this study should be acknowledged. First, according to the diagnostic guide, antihypertensive medication was one of the criteria for hypertension diagnosis, but the information was not available in the CLHLS database. In our study, we used blood pressure values and self-reported history of hypertension diagnosis to identify hypertension, which might be affected by recall bias. However, it could be confirmed that individuals diagnosed with hypertension by a physician should be prescribed antihypertensive medication (16). The address of each participant was deleted from the CLHLS database and we could not collect the precise addresses of participants from this publicly open database. However, our study estimated PM_2.5 exposure in residential counties or district addresses using satellite observations and an atmospheric chemistry database that effectively employed relatively coarse spatial resolution to estimate large-area variations in air pollution exposures (11, 50). Besides, lots of factors such as economic status, disease comorbidities, and life styles, which were dynamic and not easy to assess. Lastly, because the concentrations of other air pollutants such as SO_X, NO_X, and O₃ were not unavailable in the open database, we could not confirm the effects of PM_2.5 exposure and hypertension incidence under the influence of these air pollutants. A more comprehensive evaluation should be conducted to verify the influence of various air pollutants on hypertension in future studies.

Despite these limitations, this study further added to the present evidence of the effects of PM_2.5 exposure on hypertension among the elderly. In particular, the study provided quantitative evidence relying on a 10-year period using a large population-based cohort that included a nationwide representative sample of a specific age group (≥80 years) in a developing country. The results can be used to inform public policy to address the urgent need for more rigorous environment-related policies for air pollution abatement and make efforts to protect vulnerable elderly away from such harmful exposure.

This study highlighted the effects of long-term exposure to PM_2.5, which was associated with hypertension incidence among the elderly, and indicated that there was no safe threshold for the association between PM_2.5 and risk of hypertension. Furthermore, this relationship was greater in the elderly who were without pension, living in rural areas, and residing in central/western China. Reduction approaches and policies for PM_2.5 should be developed to reduce the incidence of hypertension and improve the quality of life of the elderly.

The CLHLS questionnaires are available at https://sites.duke.edu/centerforaging/programs/chinese-longitudinal-healthy-longevity-survey-clhls/. The full datasets used in this analysis are available from the corresponding author upon reasonable request.

The studies involving human participants were reviewed and approved by the Research Ethics Committees of Duke University and Peking University approved the protocol for each wave of the CLHLS, which is the data source of this study. The survey respondents gave informed consent before participating. The patients/participants provided their written informed consent to participate in this study.

ZW and ZF: data cleaning. ZW, ZF, CW, and LL: conceptualization and visualization. ZW, CW, ZF, and WW: data curation, writing, original draft preparation, methodology, software, and reviewing. ZW, ZF, CW, LL, and WW: supervision. All authors contributed to the article and approved the submitted version.

The research was financially supported by National Natural Science Foundation of China (Grant Nos. U1911204 and 51861125203), National Key R&D Program of China (2017YFC0405900), and the Project for Creative Research from Guangdong Water Resources Department (Grant Nos. 2018, 2020).

The views expressed in this article belong to the authors and are not to the official position of any other institution or funder.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.