Fann Neal, Zanobetti Antonella, Mork Daniel, Steinhardt William, Rappold Ana G
U.S. EPA Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina.
Harvard T.H. Chan School of Public Health, Boston, Massachusetts.
Environ Epidemiol. 2024 Jan 12;8(1):e285. doi: 10.1097/EE9.0000000000000285. eCollection 2024 Feb.
Fine particle pollution is a well-established risk to human health. Observational epidemiology generally treats events as though they are independent of one another and so do not examine the role air pollution may play in promoting the progression of disease. Multistate survival models account for the complex pathway of disease to death. We employ a multistate survival model to characterize the role of chronic exposure to PM in affecting the rate at which Medicare beneficiaries transition to first hospitalization for cardiovascular disease and then subsequently death. We use an open cohort of Medicare beneficiaries and PM concentrations estimated with photochemical model predictions, satellite-based observations, land-use data, and meteorological variables. The multistate model included three transitions: (1) entry to cardiovascular hospital admission; (2) entry to death; and (3) cardiovascular hospital admission to death. The transition intensity was modeled using a Cox proportional hazards model. For a 1 µg/m increase in annual mean PM, we estimate a nationally pooled hazard ratio of 1.022 (95% confidence interval [CI] = 1.018, 1.025) for the transition from entry to first cardiovascular hospital admission; 1.054 (95% CI = 1.039, 1.068) for the transition from entry to death; 1.036 (95% CI = 1.027, 1.044) for the transition from first cardiovascular hospital admission to death. The hazard ratios exhibited some heterogeneity within each of nine climatological regions and for each of the three transitions. We find evidence for the role of PM in both promoting chronic illness and increasing the subsequent risk of death.
细颗粒物污染对人类健康构成了确凿的风险。观察性流行病学通常将事件视为相互独立的,因此没有考察空气污染在疾病进展过程中可能发挥的作用。多状态生存模型考虑了疾病发展至死亡的复杂路径。我们采用多状态生存模型来描述长期暴露于细颗粒物对医疗保险受益人群过渡到首次因心血管疾病住院以及随后死亡的速率的影响。我们使用了一个开放队列的医疗保险受益人以及通过光化学模型预测、卫星观测、土地利用数据和气象变量估算的细颗粒物浓度。多状态模型包括三个转变:(1)进入心血管疾病住院;(2)进入死亡;(3)心血管疾病住院至死亡。转变强度使用Cox比例风险模型进行建模。对于年均细颗粒物每增加1微克/立方米,我们估计从进入状态到首次心血管疾病住院的全国汇总风险比为1.022(95%置信区间[CI]=1.018,1.025);从进入状态到死亡的风险比为1.054(95%CI = 1.039,1.068);从首次心血管疾病住院到死亡的风险比为1.036(95%CI = 1.027,1.044)。在九个气候区域中的每一个区域以及三个转变中的每一个转变中,风险比都表现出一定的异质性。我们发现有证据表明细颗粒物在促进慢性病和增加随后的死亡风险方面都发挥了作用。