1. Introduction
Ambient air pollution continues to be a dire human health threat, contributing to premature deaths in different parts of the world, more especially in the urban areas of developed and developing countries, reflecting the effects of increased urbanization, rapid industrialization, intensified use of fossil fuels, and surge in traffic volumes. Ambient air pollution comprising a complex mixture of both particles and gases remains a significant public health concern that continues to be associated with both acute and chronic health effects globally, with developing countries being the most affected [
1,
2]. Furthermore, exposure to ambient air pollutants has for decades been reported as a cause of both circulatory and respiratory mortality. Various studies have reviewed the literature on the epidemiological evidence of mortality (circulatory and respiratory) associated with short-term exposure to ambient air pollutants in China, where they highlighted that climatic factors played a critical role in the association between air pollution and mortality [
3,
4].
Climate change has the ability to influence air quality and the vice-versa, as climatic factors such as temperature, barometric pressure, wind speed, and relative humidity are known to have a large impact on air pollutants’ concentration, chemistry mixture, and how far they can be dispersed or deposited. For example, an extreme change in temperature has the ability to influence the body’s ability to regulate the internal temperature in those exposed, thereby possibly resulting in heatstroke, hypothermia, heat exhaustion, hyperthermia, frostbite, and heat cramps, all of which can result in hospitalization or even death in people with underlying illnesses such as asthma which can be triggered or worsened by poor air quality. Hence, climate change has been labeled a dire and major public health concern, with its adverse effects evidently noted in every region of the world. Subsequently, the relationship between ambient temperature and daily mortality has been extensively observed with the effect of temperature on mortality said to be influenced by various factors, including, air pollution levels, geographic location, and weather patterns, among many others [
5]. Moreover, climate change (through climatic factors) and air pollution can both affect and influence the health outcomes of those exposed (directly and indirectly). Analyses conducted in Canada, Europe, China, and the United States have provided evidence of the coherence and plausibility of these associations between ambient air pollution and climate factors, suggesting a need for more studies in developing countries to ascertain the effect of ambient air pollution and climate factors on health [
6,
7]. However, the impact of ambient air quality and climate change on human health remains complex as there are a number of factors that come into play.
While the inter-relationship between air pollution and climate change is widely known and well documented, the recent surge in climate change has become an urgent call for concern prompting a need to understand the effect of interaction between ambient air pollution and climatic factors on health and mortality. Although both ambient air pollution and climatic factors have been closely associated with daily mortality; the effect of their interaction remains largely unknown even though first investigated in 1972 [
8]. Globally, recent evidence suggests that the interaction of air pollution and climate change can significantly affect health and mortality. For instance, in Wuhan, Shanghai, and Tianjin, studies showed that extreme temperatures and relative humidity exacerbated the effect of air pollution on health [
9,
10,
11]. Despite the growing prevalence of literature on short-term exposure to ambient air pollutants and its association with cardio-respiratory mortality in China, only megacities such as Beijing, Shanghai and Guangzhou are regularly focused on, hence, leaving a challenge in understanding the extent of the association in China [
3,
4]. Moreover, although there has been an increase in epidemiologic evidence of the interaction between ambient air pollution and climatic factors in recent times, studies focusing on the effect on circulatory and respiratory mortality in China (particularly in non-metropolises) remain scarce, resulting in many unanswered questions in a country with a huge burden of air pollution in the wake of climate change. With recent urbanization and mass industrialization, air pollution levels in many Chinese cities remain significantly high, exceeding the upper limits stated by both the World Health Organization guidelines and the Chinese National Ambient Air Quality Standards. Additionally, considering the acceleration of climate change and the recent exponential increase in non-communicable diseases (particularly circulatory and respiratory) in China, the interaction effect of air pollution and climate change on mortality warrants investigation.
4. Discussion
Climate factors such as temperature and relative humidity play a pivotal role in determining the distribution, patterns, and concentrations of air pollution in our environment at any given time. Thus, climate factors and air pollution are closely associated with human health. However, these parameters are influenced by environmental and geographic factors; as such, their actions vary to a certain degree, thereby creating a need to fully understand their associations, interactions, and how they are likely to contribute to health and mortality in different regions. This study explored the basic and complex relationship between meteorological parameters and air pollution, in a bid to understand their effect on circulatory and respiratory mortality. The preliminary analysis (descriptive statistics and correlation) revealed basic but vital information about each variable. The descriptive statistics highlighted PM
2.5 as the only air pollutant that exceeded both its daily and yearly limits, hence revealing the severity of PM
2.5 emission in the city; a finding that has also been established by other ecological studies [
18,
19]. For the meteorological parameters, the averages suggest that the study period was relatively cool and moist; nonetheless, this does not imply mortality (circulatory or respiratory) was associated with cold temperatures or moist atmospheric condition only. It is important to note that both cold and hot temperatures have differing effects on circulatory and respiratory mortality. As seen in
Figure 2, increased relative risk was observed at both high and low temperatures for each mortality, while at low and average percentages of relative humidity, increased relative risk was observed for both types of mortality, although the risk associated with low temperature and relative humidity surpasses the risk associated with higher temperature and relative humidity. The effect of cold temperature has been known to increase cardiovascular strain in healthy individuals via physiological reactions targeted to maintain heat balance. These reactions may be exacerbated in individuals with cardiovascular conditions involving altered nervous system, cardiac and circulatory function [
20,
21]. Low relative humidity, on the other hand, has been associated with increased loss of water from the body, through the skin as well as the mucus membrane (in an attempt to moisten air flowing to the lungs). The less humid the air, the more moisture the body attempts to put into the air flowing to the lungs, subsequently resulting in a strain on the respiratory tract and system [
22]. The average AQI concentration in Xi’an was 105 µg·m
−3, which is within the slightly polluted index range. This pollution level has been considered a health threat for immuno-compromised people, suggesting people with underlying circulatory and respiratory diseases were at an increased risk of mortality when exposed to such air pollution concentrations. The positive correlations observed indicate that increment per unit of one variable was directly proportional to increment of the other variable being correlated (revealing if collinearity exists between said variables). For instance, the significant positive correlation observed between O
3 and temperature suggests that, as temperature increases, O
3 concentration levels also increase. This was shown to be accurate as ozone levels were noted to be high during the hot seasons (graphically displayed in our previous study) [
13,
16]. On the other hand, a negative correlation indicates two variables are inversely proportional; therefore, the negative relationship between circulatory mortality and temperature denotes that, as temperature increases, circulatory mortality decreases. However, this will only be completely correct if the relationship is a perfectly linear one; however, all bivariate relationships are not linear. Hence, the correlation analysis was used to reveal relationships (not associations) between variables, which was advantageous in structuring the model of the study.
Evidence suggests that interaction between air pollution and climatic factors exist and may have an effect on mortality, a revelation which is a major concern considering accelerated global climate change, a recent sharp increase in non-communicable diseases, and exponentially increased levels of air pollution, particularly in China. Patterns of temperature and mortality have been presented by various studies with U, V, and J curves [
12,
23,
24]. The cumulative effects of temperature in Xi’an presented a typical U-shaped exposure-response curve in relation to circulatory mortality, whereas respiratory mortality was a reversed J-shape. This suggested that circulatory mortality was mostly higher during extreme cold and warm days, an observation that has been previously reported by Tian et al. in their Beijing study, where they noted that cold effects can last for weeks in comparison to hot effects which only last for a few days [
25]. With regards to respiratory mortality, the reversed J-shaped exposure-response association with temperature suggest that colder periods were responsible for a substantial fraction of respiratory mortality [
26]. This outcome was consistent with Gasparrini et al., whose findings stated that cold temperatures were responsible for a substantial percentage of deaths compared to hot days [
27]. On the contrary, the outcome was the opposite of that of Pinheiro et al. who reported elevated mortality at high temperatures (i.e., J-shaped association curve) [
28]. Various studies have noted cold and warm temperatures as having adverse health effects that could potentially lead to fatality in people with different diseases (including circulatory and respiratory diseases) [
25,
29]. Generally, relative humidity at high levels is known to diminish the body’s ability and effectiveness in transporting metabolic heat; while at low levels, relative humidity leads to dehydration, all of which could be fatal [
30]. Notwithstanding this physiological and biological importance, relative humidity remains rarely examined independently in health and environmental factors association studies, except as a confounding variable. Consequently, the lack of studies investigating the role of relative humidity on health can potentially underestimate its effect, particularly with the continued rise of climate change. In the current study, relative humidity was closely associated with both circulatory and respiratory mortality at lower ranges, suggesting that mortality was high during less humid days. The findings were in agreement with Barreca, who reported that the effects of relative humidity on mortality were more common in cold counties in the United States than in hot counties [
31]. With the respiratory system easily exposed to the outside environment, in Taiwan relative humidity at lower levels was closely associated with chronic obstructive pulmonary disease exacerbation during cold winter times, a potentially fatal scenario [
32]. With regards to AQI, a J-shaped association was noted between AQI and circulatory mortality signifying that higher AQI was directly proportional to mortality [
33]. Similarly, an increase in AQI was associated with an increase in respiratory mortality; however, an increase in AQI beyond 350 was shown to be associated with a lesser risk of respiratory mortality. The exact reason for this was not established. However, the fact that ecological-health-related relationships do not always follow a linear trend explains a great deal. Most such relationships are non-linear; hence they often show a varying pattern of exposure–response rather than a straight line [
34].
Generally, the effects of interactions between AQI and climatic factors on circulatory mortality have a relationship. In relation to circulatory mortality, the RR (95% CI) for the interaction terms for temperature and humidity (T = 1, AQI = 1 and RH = 1, AQI = 1) were observed at 0.843(0.782, 0.909) and 1.126(1.037, 1.223), respectively. Additionally, for temperature, the IRR and RERI for the interaction in relation to circulatory mortality were recorded as 0.973(0.969, 0.977) and −0.055(−0.059, −0.048), respectively; while the IRR and RERI for relative humidity were 1.098(1.011, 1.072) and 0.088(0.081, 0.107), respectively. With regards to respiratory mortality, the RR (95% CI) for interaction terms for both for temperature and humidity (T = 1, AQI = 1 and RH = 1, AQI = 1) were 0.784(0.613, 1.003) and 1.069(0.913, 1.253), respectively. Furthermore, the IRR and RERI for the interaction in relation for respiratory mortality were 0.805(0.722, 0.896) and −0.235(−0.269, −0.163) for temperature, and 1.008(0.965, 1.051) and −0.031(−0.088, 0.025) for relative humidity. These results suggest that temperature did not interact with AQI to improve or influence the individual effect of AQI (i.e., an antagonistic relationship). However, taking into consideration that the air quality index represents the concentration level of six criteria air pollutants (some of which do not typically correlate positively with temperature), the observation of temperature’s interaction with individual air pollutants was paramount. A number of studies have highlighted the outcome of PM
2.5 and temperature’s interaction on human health; Kioumourtzoglou et al. observed a higher association between long-term PM
2.5 exposure and mortality in warmer cities; Imaizumi et al. noted a coexistence of low temperature and high PM
2.5 was associated with a 2.3-fold increased likelihood of morning hypertension. [
35,
36]. Additionally, analyzing the main effects and interaction effects of individual air pollutants gave a holistic view of air pollution’s effect on respiratory and circulatory mortality. Based on the findings of this study, the interactions between climatic factors and air pollutants are proven to contribute relevantly to changes in both circulatory and respiratory mortality. However, from the two climatic parameters observed, temperature interactions with individual air pollutants were more positive in leading to an increase in mortality, as compared to relative humidity. Therefore, the risk associated with the temperature at its minimum (−8 °C) and maximum (34 °C), were analyzed in order to establish the cold and hot effect on both mortality categories. The results revealed that both cold and hot temperatures were associated with an increased risk of circulatory and respiratory mortality [
37,
38,
39]. At its minimum, the effect of temperature was statistically significant for both circulatory and respiratory mortality with estimates 1.745(0.547, 5.567) and 10.652(0.987, 11.501) respectively, while at its maximum the effect of temperature was statistically significant for circulatory mortality only.
In comparing the risk associated with cold effects on both types of mortality, the estimates clearly show there is a higher risk of respiratory mortality due to very low temperature than circulatory mortality [
40]. While hot and cold effects are evidently associated with circulatory mortality, the risk associated with extreme cold is slightly higher. These findings are supported by the fact that the respiratory organs are usually burdened by illnesses during cold seasons and the circulatory organs are mostly stressed and overworked during the hot seasons [
41,
42,
43]. Given that these temperatures do not last for only a single day but might stretch out over several days and their effects might not be observed on the initial day of exposure, their delayed effects were analyzed over 30 lag days [
44]. The results of the current study show that both cold and hot effects on circulatory and respiratory mortality were harvested at different lags [
44,
45]. The cold effect of temperature was significantly associated with the risk of circulatory and respiratory mortality at lag 7 [
40], while the hot effect was significantly associated with risk of circulatory mortality at lag 0 and 21 and with respiratory mortality at lag 7 [
46]. Studies in and out of China have related cold and hot effects of temperature to circulatory and respiratory mortality, with most revealing somewhat similar findings to the ones presented herein [
47,
48,
49,
50,
51]. A study in Serbia reported heat-related cardio-respiratory mortality was maximum on the same day of exposure or lags 1–3, while cold spells were distributed over lag 3–6 [
40]. Similar to our study, hot effects were observed on the current day of exposure as well as at lag 7 for circulatory mortality, while the cold effect was observed at lag 7 for respiratory mortality. They also reported the highest relative risk for mortality noted in their study was attributed to the cold days, which is also in line with the relative risk associated with respiratory mortality in this study [
40]. Gasparrini et al. performed a multi-city study in several Asian and western countries and showed the effect of cold and hot temperatures on mortality, with cold temperature having the most attributed temperature-related deaths, which corresponds to the findings of our study [
28,
52,
53].
The regression analysis carried out generated results revealing the main effects and interaction effect of climatic factors and AQI. The results from the analysis prove that climatic factors and AQI are relevant variables in predicting expected changes in both circulatory and respiratory mortality. The regression coefficients showed each variable and the interaction term contributed to changes in the outcome of interest (circulatory and respiratory mortality). They also indicate the direction of the contribution, that is positive or negative. For both mortality categories, AQI had a positive coefficient, while the climatic factors had negative coefficients. The interaction between climatic factors and AQI had a positive coefficient for circulatory mortality only. It is important to note that a negative coefficient of a variable does not equate to the non-relevance of the variable. It only reveals the less contributory factor to the outcome. Looking at the main effects of each variable, it is safe to say that a unit increase of temperature (°C) and relative humidity (%) would increase circulatory and respiratory mortality by −0.216% and −0.164%, respectively. In other words, per unit increase of this climatic factor both mortalities decrease by the percentage given above. This finding corresponds accordingly with the correlation analysis, which shows both climatic factors were not directly proportional with both mortalities. The regression coefficient and correlation also suggest that, as the climatic factor decreases, mortality increases. The positive coefficients of AQI indicate that the variable contributes 0.005% and 0.062% to circulatory and respiratory mortality per unit (µg·m−3) increase. The interaction term which had a positive coefficient for circulatory mortality only indicates that the interaction between climatic factor and AQI was synergistic, given that the interaction effect of the variables on circulatory mortality was greater than the main effects of each variable on circulatory mortality. In regard to respiratory mortality, the interaction effect was antagonistic especially in consideration of AQI’s main effect. This is so since the main effect of AQI was positive, contributing 0.062% to respiratory mortality; however, interaction with climatic factors changed the main effect of AQI, resulting in a negative contribution of −0.019%. All variables and interactions were shown to have contributed to circulatory and respiratory mortality with the exception of (Humidity × Ozone). The interaction term—(Humidity × Ozone) indicates that no contribution (negative or positive) was noted from the pair. However, it is important to note that the main (sole) effect of O3 was observed to contribute 0.0005% to respiratory mortality for every 1 µg.m−3 increase, suggesting that humidity interacted antagonistically with O3 in relation to respiratory mortality. On the other hand, effect of O3 interaction with temperature on both types of mortality was synergistic, contributing 0.070% and 0.018% change in circulatory and respiratory mortality, respectively, per unit increase. PM2.5 main effect and interaction effect with temperature contributed positively to changes in both types of mortality, while its interaction with humidity contributed positively to changes only in respiratory mortality. The main effect of SO2 was positive for respiratory mortality only; its interaction with temperature and humidity was however positive for circulatory mortality. This reveals that the interaction of climatic factors and SO2 on circulatory mortality leads to increased changes in mortality, rather than the sole effect of SO2. In all the air pollutants observed, SO2 contributed the highest to respiratory mortality with 0.125% in terms of the main effect, while PM2.5 contributed the highest to circulatory mortality with 0.036%. With regards to interaction effect, temperature interaction with O3 contributed the most to the increase in circulatory mortality, while humidity interaction with PM2.5 contributed the most to respiratory mortality.
Strengths and Limitations
To the best of our knowledge, this study is the first of its kind in China to explore the interaction of air pollution and climatic factors on cardiovascular and respiratory mortality. It highlights major contributors in the relationship between air pollution and cardiovascular and respiratory mortality in Xi’an. Results for both main and combined effects strengthen the awareness that climatic conditions play a vital role in exacerbating or limiting the effect of air pollution on human health. However, the study was limited in one aspect; it lacked the stratification of mortality data into common subgroups.