The government of China reported that the average PM2.5 concentration in Beijing, Tianjin, Hebei, and surrounding areas decreased by 14.6% year-on-year in 2014, and the average PM2.5 concentration in Beijing, Tianjin, and Hebei decreased by 39.6% in 2017 compared to that in 2013[18, 27]. During 2009–2010 and 2012–2013, the mortality of ICH in Tianjin exhibited an upward trend. However, the mortality of ICH among residents in Tianjin immediately declined significantly and has been declining yearly since the air pollution control policies began to intervene in 2014.
According to the 12-year time series of death cases of ICH among Tianjin residents fitted by the ARIMA model, we found that the secular trend change caused by the intervention of air pollution control policy on ICH deaths in June 2014 was statistically significant. At that time, the " Key Work of Joint Prevention and Control for Air Pollution in Beijing-Tianjin-Hebei and Surrounding Areas in 2014 " was officially published, and several policy documents and safeguard measures favorable to the prevention and control of regional air pollution were formulated and introduced, meanwhile unified actions were carried out in Beijing-Tianjin-Hebei and surrounding regions to prevent and control air pollution. Immediate changes in ICH deaths among Tianjin residents in October 2017, December 2017, and March 2018 were statistically significant. It may be due to the intensified assessment inspection and efforts to complete the assessment target at the beginning or the end of the policy implementation. Therefore, the air pollution in Tianjin improved at those time points, leading to significant decreases in ICH deaths.
Stratified analysis by gender showed gender differences in the effect of the intervention on ICH deaths in Tianjin, with a statistically significant secular trend change in ICH deaths for females only in June 2014 and statistically significant immediate changes for females only in December 2017 and March 2018. After excluding the sequelae of ICH (the ICD-10 code is I69.1), analysis of acute ICH deaths in Tianjin revealed no statistically significant change in immediate and secular trend changes in acute ICH deaths for all residents or females and only statistically significant immediate changes for males in October 2017. These results may indicate that surviving ICH patients and females are more sensitive to the protective effects of air pollution control policies.
A UK cohort study accounting for a wide range of potential confounders showed that the survival rates of stroke patients would have declined if they lived in areas with higher levels of air pollution[28]. Another large population-based cohort study showed that patients with hemorrhagic stroke are at high risk for secondary cardiovascular disease. Compared with ischemic stroke patients, cardiovascular and all-cause mortality risk in patients with hemorrhagic stroke increased significantly during follow-up[29]. Another analysis of four population-based studies concluded similarly that ICH was associated with an approximately 2-fold increased risk of arterial ischemic event, ischemic stroke and myocardial infarction[30].
Many studies consistently show that women have a worse prognosis after stroke than men[31–33]. Although stroke morbidity and mortality rates are higher in men at specific ages than in women, more women are affected by strokes because women live longer than men, and stroke incidence is significantly higher in the oldest age group[34, 35]. In addition, stroke-related outcomes such as disability and quality of life (QOL) are consistently worse among women than men[35].
Several related studies have shown similar results to our study, suggesting that air pollution increases the risk of ICH death. It was found that short-term exposure to NO2 was associated with increased hospital admissions for ICH in a study involving 14 major cities in China[36]. There was also a positive association between short-term exposure to PM2.5 and mortality from hemorrhagic stroke, with each 10µg/m3 increase in PM2.5 associated with a 0.37% (95%CI, 0.07–0.67%) increase in mortality from hemorrhagic stroke according to a study in Beijing, China[37]. Moreover, a study in Shanghai, China, concluded similarly that the incidence of fatal ICH was associated with PM2.5 exposure[38].In addition, short-term exposure to ambient So2 could increase the risk of hospitalization for hemorrhagic stroke under a study in Guangzhou, China. Furthermore, the effect that SO2 had on the risk of hemorrhagic stroke reached the maximum value on lag1 day, with a percentage change of 1.55% (95%CI, 0.02–3.11%) per 10 µg/m3[39]. It was suggested that the risk of ICH increased after short-term exposure to ozone in a time-stratified bidirectional case-crossover analyses conducted in the Greater Boston area of the United States[40]. A study in Korea also found a correlation between ozone exposure and subarachnoid hemorrhage as well as a positive association between PM10 and the incidence of ICH[41]. However, some studies in developed Western countries found no significant association between air pollution and ICH deaths[42, 43]. It is probably due to the differences in the composition of air pollutants, pollution levels, meteorological factors, population susceptibility, and other factors between developed and developing countries[44].
Exposure to air pollution is an important risk factor for cardiovascular disease and leads to an increased risk of hemorrhagic stroke by biologically plausible mechanisms. Some pathophysiological alterations caused by air pollution may be related to ICH, such as arterial vasoconstriction, increased blood pressure, and increased vulnerability to cerebral vascular rupture due to air pollution-induced endothelial dysfunction[45]. However, further research is required.
In Fig. 1, the difference between the number of ICH and acute ICH deaths in Tianjin from 2009 to 2020 showed an increasing trend, indicating a decrease in acute deaths among all ICH deaths. It was probably due to the timelier treatment received by residents after the onset of ICH, which is associated with the improvement of economic development, the increase in medical resources, the promotion and education of prevention, and the adjustment of health insurance policies[46]. In recent years, Tianjin has established 22 stroke centers, which have played an essential role in early diagnosis of stroke, shortening the time to treatment and reducing stroke deaths.
In addition to the interventions of air pollution control policy, other interventions that led to the reduction of ICH deaths in Tianjin may include the publicity and education of healthy lifestyles and the implementation of community health management for hypertensive patients since 2008, and the enactment of the Tianjin Act of Tobacco Control in 2012[12, 47]. However, these preventive and control measures implemented did not coincide with the time points at which ICH deaths declined in Tianjin, suggesting that the decline is more likely related to the interventions of air pollution control policies.
Extensive literatures have reported the association between air pollutants and stroke. However, few studies have investigated the effects of air pollution control policies on stroke and considered the types of stroke. The interventions of the Chinese government to reduce air pollution offered a rare opportunity for quasi-experimental studies. This study was carried out in the area with higher pollution levels than North America and Europe, where air pollution levels are typically low. Thus the results may be more beneficial for other low and middle-income nations with higher levels of air pollution.
The interrupted time series design has been used extensively in evaluating policies. The assessments are based on the observation of a relatively fixed population over time and therefore are not affected by inter-group differences such as selection bias or unmeasured confounders. In addition, ARIMA models control the effects of within-group characteristics that slowly change long-term over time by modeling potential trends, such as rising economic levels and aging[48]. Another strength of this study is the data obtained from The Tianjin All Causes of Death Surveillance System. The dataset involves a relatively fixed population and covers a long period (144 months in total). Using this population-based dataset, we were able to offer greater statistical power to detect secular trends and assess the health impact.
There are also some potential limitations in this study. First, as an ecological study, it does not reflect the actual individual exposure levels. At the same time, there is a lack of comparison with other cities during the same period. Since corresponding air pollution control policies were introduced throughout China during the study period, it is difficult to find a comparable blank control group. In addition, there may have been intervening factors that could directly or indirectly affect the outcome variables during the study period, so a clear causal relationship between the reduction in ICH deaths and air pollution control measures cannot be established[48].