研究背景：慢性非传染性疾病（Non-Communicable Diseases，NCDs）已成为全球疾病负担的核心组成部分。我国缺血性卒中（Ischemic Stroke，IS）、慢性阻塞性肺疾病（Chronic Obstructive Pulmonary Disease，COPD）和糖尿病（Diabetes Mellitus，DM）等主要NCDs的发病率和死亡率持续攀升。研究表明，大气可吸入细颗粒物（Particulate Matter with an aerodynamic diameter ≤ 2.5 µm，PM₂.₅）及极端气象条件是诱发上述疾病急性加重和死亡的重要外源性危险因素。然而，现有研究仍存在三方面缺口：（1）PM₂.₅不同化学组分的特异性毒性机制尚不明确；（2）温度与污染物的协同效应缺乏系统性流行病学验证；（3）脆弱人群的识别缺乏系统性数据支撑。因此，本研究旨在系统评估北京市PM₂.₅主要组分与气象因素的复合暴露特征，量化其对IS、COPD和DM死亡风险的短期影响及交互作用，探索人群脆弱性，为精准干预和健康预警提供科学依据。

研究方法：本研究整合2008-2011年北京市逐日死亡登记数据、PM₂.₅主要化学组分浓度〔硫酸盐（SO₄²⁻）、硝酸盐（NO₃⁻）、铵盐（NH₄⁺）、有机质（Organic Matter，OM）、黑碳（Black Carbon，BC）〕，以及国家气象站记录的日均气温和相对湿度，并将二氧化氮（NO₂）、二氧化硫（SO₂）和可吸入颗粒物（PM₁₀）作为协变量纳入模型控制。数据经多重插补处理缺失后，采用准泊松回归框架下的分布滞后非线性模型（Distributed Lag Nonlinear Model，DLNM），通过交叉基函数捕捉非线性暴露–反应关系及7日内滞后效应，设定PM₂.₅组分与气温交互项。同时，按性别与年龄（65岁为界）开展分层分析，并通过调整滞后窗（0-4天、0-14天）与样条自由度进行敏感性检验，评估结果稳健性。

研究结果：研究期间，北京市SO₄²⁻、NH₄⁺、NO₃⁻、OM、BC日均浓度分别为11.24 μg/m³、8.37 μg/m³、12.00 μg/m³、17.34 μg/m³、3.32 μg/m³。OM每增加1个四分位间距（Inter Quartile Range，IQR，12.33 μg/m³），在滞后第2日和第6日明显增加DM [（相对危险度（Relative Risk，RR）RR=1.038，95%可信区间（Confidence Interval，CI）：1.005-1.071）]和COPD（RR=1.013，95% CI：1.001-1.026）死亡风险；BC每增加1个IQR（2.06 μg/m³），在滞后第0日与第6日明显增加COPD死亡风险（RR=1.228，95% CI：1.017-1.482；RR=1.059，95% CI：1.014-1.106）。SO₄^2-（IQR=7.53 μg/m³）和NH₄⁺（IQR=7.13 μg/m³）的7日累积暴露与IS死亡呈现正相关（RR=1.085，1.010-1.167；RR=1.083，95% CI：1.003-1.169）。

在气象因素方面，气温与IS、COPD和DM死亡风险均呈“U”型或“J”型非线性关联。其中，IS死亡风险在滞后第6日随每增加1个IQR（21.8℃）而升高（RR=1.052，95% CI：1.009-1.097），COPD在滞后第2日达到最大效应（RR=1.158，95% CI：1.019-1.315）。相对湿度对DM死亡具有明显的短期累积效应，在滞后第0-3日内每增加1个IQR（33%），风险分别为1.042（95% CI：1.001-1.084）、1.064（95% CI：1.023-1.106）和1.059（95% CI：1.014-1.106）。交互分析结果显示，温度对PM₂.₅各组分的急性毒性作用未表现出有统计学意义的调节效应。

分层分析发现，不同年龄与性别人群在暴露响应上存在差异。相比于65岁及以上人群，低于65岁人群对SO₄²⁻、NH₄⁺、BC及温度变化更敏感；男性在PM₂.₅暴露下IS死亡风险更高；女性则对低温与高湿条件下COPD与DM死亡更为易感。

研究结论：本研究系统揭示了PM₂.₅组分与气象因素对IS、COPD和DM死亡风险的短期影响特征。SO₄²⁻与NH₄⁺是IS死亡的关键污染因子，OM与BC对DM和COPD具有明显的急性毒性；气温对三类疾病死亡均呈非线性强化效应，相对湿度则对DM表现出明显的短期滞后作用。尽管未发现PM₂.₅组分与温度的交互作用，但人群脆弱性分析提示65岁以下人群和女性应作为重点防护对象。研究结果为推进污染组分精准治理、构建污染–气象协同健康预警体系提供了重要的实证支持与政策依据。

Background

Non-communicable diseases (NCDs) have become a central component of the global disease burden. In China, the incidence and mortality of major NCDs-including ischemic stroke (IS), chronic obstructive pulmonary disease (COPD), and diabetes mellitus (DM)-continue to increase. While growing evidence suggests that fine particulate matter with an aerodynamic diameter ≤2.5 µm (PM₂.₅) and extreme weather conditions are key environmental triggers for acute exacerbation and mortality from these diseases, knowledge gaps remain in three areas: (1) the component-specific toxicity of PM₂.₅ is poorly characterized; (2) the synergistic effects of temperature and air pollutants lack robust epidemiological validation; and (3) population vulnerability is underexplored due to limited stratified data. This study aims to systematically evaluate the compound exposure of PM₂.₅ components and meteorological factors in Beijing and quantify their short-term effects and interactions on mortality from IS, COPD, and DM.

Methods

Daily mortality records in Beijing from 2008 to 2011 were linked with measurements of PM₂.₅ components-including sulfate (SO₄²⁻), nitrate (NO₃⁻), ammonium (NH₄⁺), organic matter (OM), and black carbon (BC)-and daily meteorological data (temperature and relative humidity). Nitrogen dioxide (NO₂), sulfur dioxide (SO₂), and inhalable particles (PM₁₀) were included as covariates. After multiple imputation for missing data, a distributed lag nonlinear model (DLNM) under a quasi-Poisson framework was used to capture nonlinear exposure-response relationships and 7-day lagged effects via cross-basis functions. Interaction terms between PM₂.₅ components and temperature were included, and stratified analyses were conducted by sex and age (<65 vs. ≥65 years). Sensitivity analyses were performed by adjusting lag windows (0-4, 0-14 days) and spline degrees of freedom.

Results

During the study period, the average daily concentrations of SO₄²⁻, NH₄⁺, NO₃⁻, OM, and BC in Beijing were 11.24 μg/m³, 8.37 μg/m³, 12.00 μg/m³, 17.34 μg/m³, and 3.32 μg/m³, respectively.

An interquartile range (IQR) increase in OM (12.33 μg/m³) was significantly associated with elevated mortality risks from DM on lag days 2 and 6 (Relative Risk [RR] = 1.038, 95% Confidence Interval [CI]: 1.005-1.071) and from COPD (RR = 1.013, 95% CI: 1.001-1.026). An IQR increase in BC (2.06 μg/m³) was also significantly associated with increased COPD mortality on lag day 0 (RR = 1.228, 95% CI: 1.017-1.482) and lag day 6 (RR = 1.059, 95% CI: 1.014-1.106). Seven-day cumulative exposure to SO₄²⁻ (IQR = 7.53 μg/m³) and NH₄⁺ (IQR = 7.13 μg/m³) was positively associated with ischemic stroke (IS) mortality (RR = 1.085, 95% CI: 1.010-1.167; RR = 1.083, 95% CI: 1.003-1.169).

Regarding meteorological factors, temperature showed U-shaped or J-shaped nonlinear associations with mortality from IS, COPD, and DM. Specifically, IS mortality increased on lag day 6 with each IQR increase in temperature (21.8°C) (RR = 1.052, 95% CI: 1.009-1.097), while COPD mortality peaked on lag day 2 (RR = 1.158, 95% CI: 1.019-1.315). Relative humidity had a significant short-term cumulative effect on DM mortality; each IQR increase (33%) over lag days 0-3 was associated with RR values of 1.042 (95% CI: 1.001-1.084), 1.064 (95% CI: 1.023-1.106), and 1.059 (95% CI: 1.014-1.106), respectively. Interaction analysis revealed no statistically significant effect modification by temperature on the acute toxicity of PM₂.₅ components.

Stratified analyses revealed notable differences in exposure responses by age and sex. Compared with those aged 65 and older, individuals younger than 65 were more sensitive to SO₄²⁻, NH₄⁺, BC, and temperature variations. Male individuals exhibited a higher risk of IS mortality in relation to PM₂.₅ exposure, while females were more susceptible to increased mortality from COPD and DM under low-temperature and high-humidity conditions.

Conclusions

This study provides comprehensive evidence that PM₂.₅ components and meteorological factors significantly influence short-term mortality risks of IS, COPD, and DM, with component-specific and population-specific effects. SO₄²⁻ and NH₄⁺ are key risk factors for IS, while OM and BC exhibit strong acute toxicity in COPD and DM. Temperature exerts nonlinear reinforcing effects on all three diseases, and humidity has a clear lagged impact on DM. Although no significant interaction was detected between PM₂.₅ and temperature, vulnerable subgroups-particularly individuals under 65 and women-should be prioritized in intervention strategies. Findings support precision pollution control and integrated weather-pollution health alert systems in China.