Coronary heart disease (CHD), also known as coronary artery atherosclerotic heart disease, is a common cardiovascular condition primarily caused by the narrowing or blockage of coronary arteries. Between 1990 and 2019, the global prevalence of cardiovascular diseases nearly doubled, increasing from 271 million to 523 million cases, while the number of deaths rose from 12.1 million to 18.6 million (1). In addition to posing a significant threat to human health, CHD is expected to impose a substantial economic burden on healthcare systems, particularly in China, where the economic impact of CHD is more severe than in developed countries (2). CHD is a multifactorial disease, with risk factors that include age (3), gender (4), family history (5), environmental influences, genetic predispositions (6), and obesity (7). Notably, exposure to secondhand smoke, in addition to active smoking (8), significantly increases the risk of developing CHD (9). Our team's previous research also indicates a close association between birth month and coronary artery disease (CAD) risk (10), with individuals born in winter exhibiting a higher CAD risk. Furthermore, the incidence of CHD tends to increase with elevated fasting blood glucose levels (11).

Recent research increasingly indicates that air pollutants are significant risk factors for CHD-related hospitalizations and fatal events. For example, a time series analysis in Lanzhou found that each 10 μg/m³ increase in NO₂, SO₂, PM₂.₅, and PM₁₀ concentrations was associated with a rise in CHD hospitalization risk by 0.20%, 0.53%, 0.14%, and 0.03%, respectively (12). CO had the greatest impact, with each 1 mg/m³ increase raising hospitalization risk by 10.76%, peaking on the third day. In contrast, O₃ concentrations showed a protective effect, with each 10 μg/m³ increase reducing hospitalization risk by 0.09%. Additionally, elevated concentrations of PM₂.₅, PM₁₀, SO₂, CO, and NO₂ were associated with an increased risk of hospitalization for CHD in Beijing (13). These pollutants also exhibited a time-lag effect, with their impact being more pronounced under low-temperature conditions. A study in Wuhan (14) similarly demonstrated a significant positive correlation between the concentrations of PM₂.₅, PM₁₀, NO₂, and SO₂ and hospitalizations for ischemic heart disease (IHD), with the impact of air pollution being especially noticeable during the cold season and among elderly individuals (≥76y). Similarly, air pollution in coastal cities of southern China imposes a substantial burden on IHD, with SO₂ and NO₂ exerting a more pronounced influence than PM₂.₅ (15). Although studies on the relationship between air pollution and CHD have originated from diverse geographic locations and used different statistical methodologies, findings consistently highlight the close association between air pollutants and the development and progression of CHD.

Henan Province, located in central China, is one of the most populous provinces in the country, with an estimated 12–16 million heart disease patients. Cardiovascular disease-related deaths rank second nationwide, underscoring substantial public health challenges within the province. In addition to the prevalent dietary habits characterized by high fat, salt, and oil intake, the impact of air pollution on the health of CHD patients in Henan cannot be overlooked. Henan is among the provinces with the most severe air pollution in China, accounting for half of the regions within the bottom 20 rankings nationwide for air quality. According to Air Pollution Characteristics and Health Risks in Henan Province, China (16), calculations of the Health Air Quality Index (HAQI) indicate that Henan residents are consistently exposed to polluted air, with approximately 28% of the population living in air conditions harmful to health and 31% exposed to extremely unhealthy air environments. The study emphasizes that PM₂.₅ and PM₁₀ are the predominant pollutants, with air quality particularly deteriorating in winter due to coal heating and unfavorable weather conditions. Additionally, Henan has a high rate of vehicle ownership, with 25.88 million vehicles contributing significantly to air pollution through vehicle emissions. Given these factors, understanding the relationship between air pollution and CHD incidence in Henan Province is of critical importance.

Currently, research on the correlation between air pollution and CHD in Henan Province remains limited. Therefore, this study utilizes a time-stratified case-crossover design combined with DLNM to assess the short-term effects of six air pollutants (PM₂.₅, PM₁₀, NO₂, SO₂, CO, and O₃) at extreme concentrations on CHD hospitalization risk. Additionally, it investigates the interactive effects between different pollutants on CHD hospitalizations, aiming to provide new insights into the relationship between air pollution and CHD incidence in Henan Province.

The scatter plot of air pollution concentration and the distribution of CHD hospitalizations in Henan Province from November 1, 2016, to December 31, 2021 (Figure 2) shows that the number of CHD hospital admissions was higher in the colder seasons and generally showed an upward trend during the study period. O₃ concentration was higher in the summer due to the influence of temperature and sunlight duration. Meanwhile, concentirations of PM₂.₅, PM₁₀, NO₂, SO₂, and CO remain relatively stable, with higher levels generally observed in colder seasons. But a notable exception was the period around January 1, 2016, and 2017, when the concentration of pollutants (except O₃) was significantly higher, and subsequently, pollution levels were somewhat controlled. The characteristics of study population, air pollutant concentrations, and meteorological factors are presented in Table 1. During the study period, a total of 133,294 CHD hospitalizations were recorded, with males accounting for 53.72% (71,600 cases), females 43.70% (58,249 cases), and cases with missing gender data comprising 2.58% (3,445 cases). Patients aged over 65 years (≥65y) represented 56.13% (74,813 cases), while those aged 0–64 years (≤64y) made up 43.87% (58,479 cases), with only 0.0015% (2 cases) missing age data. The average concentrations of air pollutants during this period were as follows: PM₂.₅, 61.84 µg/m³; PM₁₀, 109.18 µg/m³; SO₂, 17.97 µg/m³; NO₂, 36.53 µg/m³; CO, 1.18 mg/m³; and O₃, 104.29 µg/m³. The average daily temperature was 16.06°C, and the average daily relative humidity was 61.61%. Sunshine duration data were missing for a total of 304 days from January 1 to October 31, 2021. Therefore, “sunshine duration” was not included as a covariate in the model.

We employed cumulative exposure-response curves to illustrate the impact of varying concentrations of different air pollutants on CHD hospitalizations across distinct exposure subgroups, stratified by age and gender (Figure 3). The exposure-response relationships for PM₁₀, SO₂, and NO₂ with CHD hospitalization rates exhibited a nonlinear, no-threshold pattern (P < 0.05), whereas the trend for O₃ was precisely the opposite of that observed for the other pollutants. Notably, no significant association was detected for PM₂.₅ and CO in the exposure-response curves. Subgroup analyses revealed that younger individuals and males were more susceptible to increased CHD hospitalization risks associated with PM₁₀, SO₂, and NO₂ exposure. However, across all concentration levels and subgroups, no definitive link was established between PM₂.₅ exposure and elevated CHD hospitalization risk. In contrast, a significant association between elevated CO concentrations and increased CHD hospitalization risk was observed exclusively among individuals under the age of 65, underscoring a distinct vulnerability unique to this demographic group. Figure 4 illustrates the relative risk and 95% confidence interval of CHD hospitalizations associated with extremely high and low concentrations of air pollutants in a single-pollutant model. At extremely high concentrations, NO₂, SO₂, and PM₁₀ show a significant association with increased CHD hospitalizations compared to median levels, while extremely low concentrations exhibit a protective effect. The lag effect is strongest on the 4th single day and the 7th cumulative day. Across cumulative lag days, the impact of extremely high and low concentrations of NO₂, SO₂, and PM₁₀ on CHD hospitalizations follows a linear pattern, with the cumulative effect of extremely high concentrations intensifying with extended lag periods, the maximum RR values are 1.768 (95% CI: 1.495, 2.091) for NO₂ (99th vs. median), 2.821 (95% CI: 2.314, 3.440) for SO₂ (99th vs. median), and 1.728 (95% CI: 1.440, 2.073) for PM₁₀ (99th vs. median). In contrast to other air pollutants, O₃ demonstrates a protective effect at extremely high concentrations compared to the median. No significant association is observed between either extremely high or low concentrations of CO and PM₂.₅ and CHD incidence across any lag days.

In the subgroup analysis based on age and gender, Supplementary Tables S2, S3 present the cumulative lagged effects of extreme pollutant concentrations on CHD hospitalizations in different subgroups, with single-day lagged effects detailed in Supplementary Tables S4, S5. In this study, extremely high concentrations of NO₂, SO₂, and PM₁₀ as well as extremely low concentrations of O₃, are associated with a higher relative risk in young (≤64y) and males. Specifically, for SO₂ and NO₂, subgroup analyses by age and gender revealed no significant differences in relative risk during the initial lag period (lag0–lag2); however, from lag3 onwards, a pronounced increase in risk emerged among younger males. This suggests that the detrimental effects of high concentrations of SO₂ and NO₂ on younger male populations are not immediate but rather exhibit a delayed onset, likely requiring cumulative exposure to manifest their full impact. During the early lag period (lag0–lag1), elderly females exhibited a higher risk of CHD hospitalization following PM₁₀ exposure, indicating greater susceptibility to its immediate effects. This heightened vulnerability may be attributed to their diminished immune tolerance and increased sensitivity, rendering them more prone to adverse cardiovascular outcomes in the initial phase of exposure. However, from lag2 onward, younger males began to exhibit a markedly higher risk than elderly females. A plausible explanation for this shift could be the increased likelihood of prolonged outdoor exposure among younger males, particularly those engaged in outdoor occupations. Coupled with lower awareness or adherence to protective measures, this sustained exposure may lead to a “threshold breakthrough” effect. Of particular note, SO₂ demonstrated the highest relative risk among the pollutants studied, with the risk among younger individuals exceeding that of the elderly by more than twofold.

Conversely, extremely low concentrations of NO₂, PM₁₀, and SO₂, as well as extremely high concentrations of O₃, demonstrated a protective effect against CHD hospitalizations when compared to median levels, with significant variations observed across different subgroups. Notably, the potential protective effect of O₃ was particularly pronounced among younger males throughout the entire lag period, warranting further investigation to elucidate the underlying mechanisms driving these disparities. Furthermore, extreme concentrations of PM₂.₅ did not exhibit any significant association with CHD hospitalizations across all subgroups, suggesting a comparatively weaker impact relative to other pollutants. Similarly, extreme CO concentrations showed no significant effects in gender-stratified analyses. However, high CO levels were associated with detrimental effects among younger individuals (≤64y) on specific single lag days (lag2–lag6) and over cumulative lag periods (lag04 to lag07), with the highest relative risk observed at lag07 (RR = 1.450; 95% CI: 1.225, 1.716). In contrast, among elderly individuals, high CO concentrations were linked to significant adverse effects on the day of exposure, indicating an immediate health impact in this population.

Spearman rank correlation analysis can measure the monotonic relationship between meteorological factors and air pollutants. As shown in Figure 5, the correlation between PM₁₀ and PM₂.₅ was the strongest, with a coefficient of 0.87, followed by the correlation between PM₁₀ and NO₂, which was 0.71. Notably, O₃ exhibited a negative correlation with other pollutants, while the remaining five air pollutants demonstrated positive correlations to varying extents. In particular, the correlation coefficients between PM₂.₅ and PM₁₀, NO₂, and CO, as well as between NO₂ and PM₁₀, SO₂, and between average temperature and O₃, all exceeded 0.6, indicating a relatively high level of association. Table 2 presents the interaction analysis of air pollutants on CHD hospitalizations. Significant synergistic interactions (P < 0.05, IRR >1, RERI >0) were observed between CO and the pollutants NO₂, SO₂, and PM₁₀. In interactions involving O₃, when PM₁₀ or CO was used as the reference pollutant, antagonistic effects were observed (IRR <1, RERI <0), with the RERI for CO and O₃ remaining non-significant. For PM₂.₅ and PM₁₀, or PM₁₀ and SO₂ (with PM₁₀ as the reference), only RERI was statistically significant (P < 0.05), while IRR demonstrated no notable effects. Among all pollutant pairs, SO₂ and CO exhibited the strongest interaction, with an IRR of 1.247 (95% CI: 1.178, 1.317) and an RERI of 0.198 (95% CI: 0.155, 0.241).

The detailed results of the sensitivity analysis are provided in the supplementary file (Supplementary Tables S6–S64). Adjusting the degrees of freedom (df) for temperature and humidity from 3 to 2 or 4, as well as modifying the maximum lag days from 7 days to 5, 6, 8, or 9, yields results consistent with the primary findings.

This study focuses on Henan Province and employs a single pollutant model in conjunction with a DLNM based on a case-crossover design to investigate the impact of individual air pollutants on CHD hospitalizations. The findings indicate that extremely high concentrations of NO₂, SO₂, and PM₁₀ significantly increased the risk of CHD hospitalizations when compared to median levels, while extremely high concentrations of O₃ demonstrated a protective effect. The observed associations between elevated concentrations of NO₂, SO₂, and PM₁₀ and CHD hospitalizations are largely consistent with previous studies (29–31).

However, the effects of O₃ are complex and variable. In most cardiovascular studies (32, 33), elevated O₃ concentrations have been shown to significantly increase the incidence of various cardiovascular diseases, particularly when O₃ levels exceed the World Health Organization's recommended threshold of 100 µg/m³ (34). A study in Sweden found that a 10 µg/m³ rise in the lag2 and lag7 average O₃ concentration led to a 0.7% and 2.7% increase in cardiovascular mortality, respectively (35). Similarly, a 2024 nationwide cohort study in China reported (36) that for every 10 µg/m³ increase in long-term O₃ exposure, the risk of CHD increased by 15%. Additionally, short-term O₃ exposure has been linked to an elevated risk of atrial fibrillation episodes (37). Nevertheless, not all studies have found a consistent association between O₃ and cardiovascular disease. An Australian time-stratified case-crossover study on air pollution and acute myocardial infarction (AMI) emergency department visits found no strong correlation between O₃ exposure and AMI risk (38). A 2023 meta-analysis further highlighted that (39) short-term exposure to PM₂.₅, PM₁₀, NO₂, and CO was significantly associated with increased hospitalization and mortality risk from heart failure, while O₃ demonstrated no significant short-term effects.

Interestingly, a 2024 case-crossover study (40) on acute cardiovascular events in New York City found that summer O₃ concentrations were negatively correlated with cardiovascular risk, potentially due to the chemical scavenging effect of O₃. Similarly, several other studies (12, 27, 41) have also documented an inverse association between O₃ levels and CHD incidence, these ozone-related findings are consistent with those of our study, while the underlying mechanisms remain incompletely understood. Previous research has suggested that O₃ may induce oxidative preconditioning, enhancing the activity of key antioxidant enzymes. This, in turn, reduces lipid peroxidation and protein oxidation, thereby mitigating oxidative stress-induced cardiac injury (42). Moreover, ozone preconditioning has been shown to activate the Nrf2 pathway, upregulating the expression of antioxidant proteins such as Slc7a11 and Gpx4 (43). This mechanism plays a crucial role in inhibiting ferroptosis triggered by ischemia-reperfusion injury, thereby attenuating myocardial damage associated with ischemic episodes. Additionally, O₃ facilitates the release of nitric oxide, a potent vasodilator that enhances endothelial function, promotes myocardial perfusion, and increases oxygen delivery to the heart. As a result, O₃ alleviates myocardial ischemia and reduces cardiomyocyte injury (44). Beyond these effects, ozone has demonstrated the capacity to improve hemorheological properties by reducing blood viscosity, thereby enhancing circulation, optimizing myocardial oxygenation, and facilitating cardiac metabolism (45). While these mechanistic insights provide compelling evidence supporting the potential cardioprotective effects of ozone, further research is required to fully elucidate the precise physiological pathways underlying these benefits.

Numerous prior studies have firmly established a strong association between CO, PM₂.₅, and cardiovascular risk. For instance, a Shanghai study (46) linked a 10 μg/m³ pollutant increase (lag0) to 0.68% and 0.08% rises in out-of-hospital CHD mortality for PM₂.₅ and CO, respectively. Similarly, In Hubei Province (47), acute exposure to PM₂.₅, CO, and other pollutants was linked to longer hospital stays for IHD patients. Furthermore, two meta-analyses (39, 48) revealed a significant increase in the risk of hospitalization and mortality due to heart failure following exposure to PM₂.₅ and CO, even at low concentrations (49). Additionally, elevated CO levels have been strongly associated with a heightened risk of heart failure rehospitalization in patients with unstable angina (50). Regarding PM₂.₅, an extensive body of literature has unequivocally demonstrated its role in elevating CHD risk (51–55). Long-term exposure to PM₂.₅ is particularly concerning, as every 10 μg/m³ increase in PM₂.₅ concentration has been linked to a 43% increase in total CHD risk (HR 1.43), a 45% increase in non-fatal CHD risk (HR 1.45), and a 38% increase in fatal CHD risk (HR 1.38) (56). A similar time-stratified case-crossover study conducted across nine cities in Sichuan Province further corroborated these findings (57), showing that for every 10 μg/m³ rise in PM₁₀ and PM₂.₅ concentrations, CHD hospitalization risk increased by 0.46% and 0.57%, respectively. Beyond that, PM₂.₅ exposure has also been linked to increased mortality and economic burden from ischemic heart disease (58–60), alongside a potential heightened risk of ventricular arrhythmias (61, 62). Strikingly, an umbrella review (63) integrating previous meta-analyses concluded that PM₂.₅ is significantly associated with nearly all major cardiovascular diseases, including mortality, myocardial infarction, stroke, hypertension, and atherosclerosis.

However, the non-significant associations between CO/PM₂.₅ exposure and CHD observed in this study diverge from previous epidemiological findings. Potential explanations include: First, our analysis utilized a maximum 7-day lag period, whereas PM₂.₅ or CO effects might manifest over longer lags (>10 days). Second, differences in study regions can also impact the experimental results. For example, Beijing reported null associations between SO₂, CO, or O₃ and out-of-hospital cardiac arrest (64), whereas Italian studies identified significant CO-OHCA risks (65). Similarly, PM₂.₅ and PM₁₀ exhibited stronger associations with acute cardiovascular events in Taiwan's central/southern regions than in northern/eastern areas (66). Furthermore, COVID-19 pandemic restrictions during the study period likely increased indoor residence time, potentially decoupling indoor pollutant exposures from outdoor monitoring data and introducing exposure misclassification. Finally, single-pollutant models were employed to avoid multicollinearity, though atmospheric pollutants typically co-occur, which may have led to the introduction of bias. Notably, synergistic effects between CO and NO₂/SO₂/PM₁₀ emerged in interaction analyses, suggesting CO's cardiovascular impacts may be amplified in multi-pollutant contexts through oxygen transport impairment, oxidative stress, and inflammatory pathways (67, 68).

In interaction analyses centered on PM₁₀, dual assessment via additive (RERI) and multiplicative (IRR) interaction indices constructed with DLNM revealed synergistic and antagonistic effects between CO/O₃ and PM₁₀ in additive and multiplicative models, respectively. In contrast, PM₁₀-PM₂.₅ and PM₁₀-SO₂ interactions showed statistical significance only in additive models, with IRR values remaining non-significant. From an epidemiological modeling perspective, this phenomenon—additive interaction reaching significance while multiplicative interaction remains null—suggests that PM₂.₅ and SO₂ primarily exert independent dose-additive effects on CHD risk rather than true biological synergy mediated by receptor colocalization or metabolic competition. For instance, both PM₂.₅ and PM₁₀ are potent inducers of oxidative stress, driving the excessive generation of free radicals that inflict damage upon cell membranes, DNA, and proteins, while simultaneously provoking inflammatory cascades within the body (69). However, due to its relatively larger aerodynamic diameter, PM₁₀ is more effectively cleared via mucociliary mechanisms, resulting in prolonged retention within the upper respiratory tract. This extended residence time not only amplifies opportunities for co-exposure with other airborne pollutants but also renders its cardiovascular effects more indirect in nature. In contrast, PM₂.₅, with its finer particulate size and significantly larger reactive surface area, possesses the capacity to penetrate deep into the alveolar regions of the lungs and even translocate into systemic circulation (70). Once within the bloodstream, PM₂.₅ exerts profound cardiotoxic effects through the induction of chronic inflammation (71), oxidative stress (72, 73), and autonomic dysregulation (74, 75), culminating in a heightened risk of cardiovascular morbidity. From a policy and regulatory perspective, for pollutant combinations displaying simple additive interactions, risk mitigation strategies may be effectively implemented through precise source tracing and independent emission control, enabling a linear reduction in health risks.

In this study, the Spearman correlation analysis of pollutants revealed a notably high correlation of 0.87 between PM₁₀ and PM₂.₅, while the lowest correlation was observed between SO₂ and PM₂.₅ at 0.47. The correlations among other air pollutants hovered around 0.6, indicating a moderately elevated interdependence. It is worth noting that the Spearman correlation coefficient is best suited for describing monotonic relationships (76). However, only specific pollutant pairings exhibited statistically significant interactive effects. For instance, a global multicenter study demonstrated that (77), under high exposure levels, the synergy index of PM₂.₅ and O₃ in relation to cardiovascular mortality reached 1.37, synergistically increasing the relative risk of CHD hospitalization.

Subgroup analysis further indicated that younger individuals and males exhibited heightened susceptibility to the detrimental effects of NO₂, SO₂, and PM₁₀, as well as an intriguing protective effect of O₃ against CHD risk. Previous studies have suggested several plausible explanations for this observation. Firstly, younger individuals and men tend to engage in outdoor activities more frequently or work in highly polluted environments, leading to increased exposure and inadequate protective measures, thereby rendering them more vulnerable to acute effects during pollution surges (78). Secondly, lifestyle factors such as smoking, alcohol consumption, and irregular sleep patterns prevalent among these subgroups may exacerbate the cardiovascular impact of air pollution (79). Additionally, stress-induced activation of the sympathetic nervous system may further heighten cardiovascular strain (80). Furthermore, males may possess relatively weaker antioxidant capacities, rendering them more susceptible to oxidative stress induced by air pollution. Experimental evidence has demonstrated that (81), following identical infusions of angiotensin II, oxidative stress levels in male mice were significantly higher than in their female counterparts. Given that younger individuals and men typically exhibit higher levels of physical activity and respiratory rates, their inhalation of airborne pollutants is correspondingly greater, potentially predisposing them to both short-term and long-term health risks (82). Notably, this study also revealed that high concentrations of CO predominantly increased the risk of CHD among younger individuals (≤64y), underscoring the necessity of stringent air pollution control measures—particularly targeting CO emissions—in major metropolitan areas where young populations predominantly reside.

This study focuses on six major air pollutants in the Central Plains region of China, an area whose air pollution levels are emblematic of broader global patterns. The findings of this research offer insights that can be extrapolated to other nations or regions facing similar challenges. For instance, according to the World Health Organization, the air pollution levels in Henan Province—an industrial and traffic-intensive hub—are comparable to those found in emerging economies such as Bangladesh and India, yet notably higher than those in many developed nations. In contrast, rural areas or western regions of China exhibit relatively lower pollution levels, more akin to the suburban air quality found in developed countries. China's air pollution, distinct from that of other countries, is primarily driven by coal combustion, industrial emissions, and vehicle exhaust, with air quality management relying heavily on national policies and exhibiting a clear urban-rural divide.

The limitations of this study include: (1) In matching patients' residential addresses with air quality monitoring stations, we used the “nearest neighbor principle,” directly linking patients’ addresses to the geographically closest monitoring station without setting a 10-km radius or other distance threshold for screening. This may lead to limitations in exposure assessment due to the spatial coverage of monitoring stations, especially when patients live far from the nearest station, potentially causing bias in exposure estimates. (2) We used hospital admission dates as a proxy for symptom onset, which may result in temporal misalignment between pollution exposure and CHD onset, leading to estimation bias. (3) Although we adjusted for covariates such as temperature, humidity, day of the week, and holidays in the model, data limitations prevented us from including other potential confounders (e.g., sunshine duration, socioeconomic status, healthcare accessibility, comorbidities) as covariates to control for bias, which may affect the results to some extent.

The findings demonstrated that short-term exposure to elevated concentrations of NO₂, SO₂, and PM₁₀ in Henan Province significantly increased the risk of CHD hospitalizations, with cumulative effects peaking at a 7-day lag. Conversely, higher O₃ concentrations exhibited a negative correlation with CHD risk. Subgroup analyses revealed greater susceptibility among younger populations and males to air pollutants, while PM₂.₅ showed no significant associations across subgroups and CO only elevated CHD risk in younger demographics. Some pollutants interact significantly, with CO showing additive and multiplicative synergy with NO₂, SO₂, and PM₁₀, suggesting its adverse impact on CHD is amplified in a multi-pollutant environment. Future large-scale studies are warranted to validate these pollutant interaction mechanisms and strengthen causal inferences.