Method for assessing the risk of tuberculosis under combined exposure to extreme weather and atmospheric pollution

Abstract

The application provides a method for evaluating the risk of tuberculosis under the combined exposure of extreme weather and air pollution, comprising the steps of: obtaining meteorological data, air pollution data and tuberculosis incidence data of a target area within a target period; identifying extreme weather events based on the meteorological data, identifying air pollution events based on the air pollution data, and identifying combined exposure events based on the meteorological data and the air pollution data, wherein the combined exposure events are combinations of the extreme weather events and the air pollution events that occur within a time interval less than a preset threshold and are spatially co-located; based on the spatial co-location, constructing a regression model with the number of cases in the tuberculosis incidence data as the dependent variable and whether the combined exposure event occurs as the binary explanatory variable, and evaluating the influence of the combined exposure event on the risk of tuberculosis. The application realizes high-precision quantitative evaluation of the risk of tuberculosis under the combined exposure of extreme weather and air pollution, and provides a scientific basis for early warning and precise prevention and control.

Description

Technical Field

[0001] This invention relates to the field of environmental and health technology, and in particular to a method for assessing the risk of tuberculosis incidence under combined exposure to extreme weather and air pollution. Background Technology

[0002] Pulmonary tuberculosis is a respiratory infectious disease caused by Mycobacterium tuberculosis. Its incidence is not only related to pathogen exposure but also influenced by environmental factors. Studies have shown that meteorological factors (such as extreme high and low temperatures) and air pollution (such as PM2.5) are associated with the risk of developing pulmonary tuberculosis.

[0003] However, current technologies mainly focus on the impact of single environmental factors on the incidence of tuberculosis, such as analyzing the effects of extreme weather or air pollution alone, and lack methods for assessing the combined effects of these two factors. When extreme weather and air pollution occur simultaneously or sequentially within a short period, they may produce synergistic or additive health damage effects, increasing the risk of tuberculosis, but existing methods are insufficient to quantify the risk of tuberculosis under such combined exposure.

[0004] In addition, existing studies mostly rely on limited data from ground monitoring stations, which have low spatiotemporal resolution and exposure misclassification, affecting the accuracy of risk assessment.

[0005] Therefore, there is an urgent need for a method that can accurately assess the risk of tuberculosis incidence under combined exposure to extreme weather and air pollution, so as to provide a scientific basis for early warning and precise prevention and control of tuberculosis. Summary of the Invention

[0006] The purpose of this invention is to provide a method for assessing the risk of tuberculosis incidence under combined exposure to extreme weather and air pollution, in order to solve the problems existing in the prior art.

[0007] The technical solution of this invention is: a method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution, comprising the following steps:

[0008] Acquire meteorological data, air pollution data, and tuberculosis incidence data for the target area within the target time period;

[0009] Extreme weather events are identified based on the meteorological data, air pollution events are identified based on the air pollution data, and composite exposure events are identified based on the meteorological data and the air pollution data. The composite exposure event is a combination of the extreme weather event and the air pollution event that occur at a time interval of less than a preset threshold and are spatially co-located.

[0010] Based on the spatial colocation, a regression model is constructed using the number of cases in the pulmonary tuberculosis incidence data as the dependent variable and whether the combined exposure event occurs as a binary explanatory variable to assess the impact of the combined exposure event on the risk of pulmonary tuberculosis incidence.

[0011] Preferably, the preset threshold is set based on the time series similarity analysis results of the extreme weather event and the air pollution event.

[0012] Preferably, the time series similarity analysis includes: calculating the Jaccard similarity between the extreme weather event and the air pollution event under different occurrence time intervals, and using the occurrence time interval corresponding to the highest Jaccard similarity as the preset threshold.

[0013] Preferably, the spatial co-location refers to the extreme weather event and the air pollution event occurring within the same spatial unit, which is obtained by dividing the target area into grids.

[0014] Preferably, the assessment method for the extreme weather event includes:

[0015] Based on the meteorological data, construct the daily maximum temperature sequence and the daily minimum temperature sequence within the spatial unit;

[0016] If the daily maximum temperature is greater than or equal to the first temperature threshold, it is a daytime heat event, where the first temperature threshold is the A percentile of the daily maximum temperature sequence;

[0017] If the daily maximum temperature is less than or equal to the second temperature threshold, it is a daytime cold event, where the second temperature threshold is the B percentile of the daily maximum temperature sequence.

[0018] If the daily minimum temperature is greater than or equal to the third temperature threshold, it is a nighttime heat event, where the third temperature threshold is the C percentile of the daily minimum temperature sequence;

[0019] If the daily minimum temperature is less than or equal to the fourth temperature threshold, it is considered a cold night event, where the fourth temperature threshold is the D percentile of the daily minimum temperature sequence.

[0020] Preferably, the meteorological data includes temperature data, air pressure data, and humidity data obtained by simulation using a WRF model based on FNL reanalysis data.

[0021] Preferably, the assessment method for the air pollution incident includes:

[0022] Based on the aforementioned air pollution data, daily dust-type pollutant concentration sequences, daily secondary pollutant concentration sequences, and daily combustion pollutant concentration sequences are constructed within the aforementioned spatial unit.

[0023] If the daily concentration of particulate matter is greater than or equal to the first pollutant threshold, then it is a particulate matter pollution event, where the first pollutant threshold is the E percentile of the daily particulate matter concentration sequence.

[0024] If the daily concentration of secondary pollutants is greater than or equal to the second pollutant threshold, then it is a secondary pollution event, where the second pollutant threshold is the F percentile of the daily secondary pollutant concentration sequence.

[0025] If the daily concentration of combustion pollutants is greater than or equal to the third pollutant threshold, then it is a combustion pollution event, where the third pollutant threshold is the G percentile of the daily combustion pollutant concentration sequence.

[0026] If two or more of the following conditions are met: daily concentration of particulate matter ≥ the first pollutant threshold, daily concentration of secondary pollutants ≥ the second pollutant threshold, or daily concentration of combustion pollutants ≥ the third pollutant threshold, then it is considered a cumulative pollution event.

[0027] Preferably, the air pollution data includes particulate matter PM2.5 obtained by simulation using the WRF-CMAQ model based on FNL reanalysis data and pollution source emission inventories. 2.5 and its chemical component concentration data; the particulate matter PM 2.5 The chemical composition includes crustal elements for identifying dust-type pollution events, ions for identifying secondary pollution events, and carbonaceous matter for identifying combustion-type pollution events.

[0028] Preferably, the regression model is a generalized additive model of a quasi-Poisson distribution, and its analytical expression is:

[0029] logE(N) = βZ coexpo +s(RH,k1)+s(PRS,k2)+ns(time,df)+DOW+sex+age+ε;

[0030] In the formula:

[0031] E(N) is the expected value of the number of cases N in the pulmonary tuberculosis incidence data;

[0032] Z coexpo It is a binary explanatory variable for whether the composite exposure event occurred;

[0033] β is the explanatory variable Z coexpo The corresponding regression coefficients;

[0034] s(RH,k1) and s(PRS,k2) are spline smoothing functions for the humidity data RH and the air pressure data PRS in the meteorological data, where k1 and k2 represent the degrees of freedom.

[0035] ns(time,df) is the natural cubic spline function of the time trend time, where df represents the degrees of freedom;

[0036] DOW is a dummy variable used to control for the weekday effect;

[0037] sex is a binary gender variable based on the tuberculosis incidence data;

[0038] age is an age grading variable based on the tuberculosis incidence data;

[0039] ε is the intercept term.

[0040] Preferably, the impact of the combined exposure event on the risk of developing pulmonary tuberculosis is assessed based on the relative risk (RR), which is expressed as RR = exp(β).

[0041] The beneficial effects of this invention include:

[0042] (1) For the first time, a quantitative assessment of the risk of tuberculosis under combined exposure to extreme weather and air pollution was achieved, breaking through the limitations of traditional research that only focuses on a single environmental factor. By constructing sixteen types of combined exposure events, the changing characteristics of the risk of tuberculosis under different combined exposure scenarios were fully revealed.

[0043] (2) By using Jaccard similarity analysis, the time interval threshold for judging whether a compound exposure event has occurred was scientifically determined, thus realizing the objective identification of compound exposure events and avoiding the arbitrariness of subjectively setting the time window.

[0044] (3) Using WRF and CMAQ models to simulate and obtain high spatiotemporal resolution meteorological and atmospheric pollution data at the kilometer and hour levels, compared with traditional monitoring station data, it can more accurately reflect the true exposure level of the population, effectively avoid the problem of exposure misclassification, and significantly improve the accuracy of risk assessment.

[0045] (4) The association between compound exposure and the risk of tuberculosis was quantified by a generalized additive model of quasi-Poisson distribution. The model effectively controlled the influence of multiple confounding factors such as humidity, air pressure, time trend, week effect, gender, and age, ensuring the accuracy and reliability of the assessment results.

[0046] (5) Based on the above method, the present invention can identify high-risk compound exposure types and their high-incidence spatiotemporal areas. Combined with meteorological forecast data and atmospheric pollution forecast data, it can predict the occurrence of compound exposure events in advance, providing scientific basis and decision support for public health departments to carry out early warning, early intervention and environmental risk management for key populations. Attached Figure Description

[0047] Figure 1This is a spatial distribution map showing the frequency statistics of four typical extreme weather and air pollution combined exposure events in Tianjin in this embodiment of the invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] The method for assessing the risk of tuberculosis incidence under combined exposure to extreme weather and air pollution provided by this invention specifically includes the following steps:

[0050] S1. Obtain meteorological data, air pollution data, and tuberculosis incidence data for the target area within the target time period.

[0051] Depending on the research objectives, meteorological data includes core meteorological element data required for the study, as well as optional data on other confounding meteorological factors that may affect the incidence of pulmonary tuberculosis. The core meteorological elements and confounding meteorological factors are determined according to research needs and may coexist or may only include the core meteorological elements. Air pollution data includes core air pollution element data required for the study, as well as optional data on other confounding air pollution factors that may affect the incidence of pulmonary tuberculosis. The core air pollution elements and confounding air pollution factors are determined according to research needs and may coexist or may only include the core air pollution elements. Pulmonary tuberculosis incidence data includes the number of cases counted at a preset time granularity, as well as information on the onset time, spatial location, gender, and age of cases, which can be obtained from disease control departments or medical institutions.

[0052] To obtain high spatiotemporal resolution exposure data and overcome the exposure misclassification problem caused by insufficient temporal and spatial resolution of ground monitoring stations, the meteorological data in this technical solution is obtained through meteorological model simulation, and the atmospheric pollution data is obtained through coupled simulation of meteorological model and atmospheric chemical transport model. The simulation data has preset spatial resolution (such as kilometer level or higher) and temporal resolution (such as hour level or higher), which can be flexibly set according to research needs.

[0053] To provide a concise and clear introduction to this technical solution, the following explanation is based on a specific research example. In this research example, the factors associated with the risk of tuberculosis include: temperature as the core meteorological element, humidity and air pressure as confounding meteorological factors, and PM2.5 and its chemical components as the core air pollution element. This research example uses the WRF model to simulate meteorological data and the WRF-CMAQ model to simulate air pollution data.

[0054] Taking a specific region as the target area, the spatial and temporal resolutions are first set according to the research requirements. The WRF model and WRF-CMAQ model divide the target area into grids according to the set spatial resolution, automatically generating multiple spatial units. According to the set temporal resolution, the target time period is divided into equal intervals, automatically generating multiple continuous time units. For example, if the time resolution is 1 hour, hour-level time units are generated; if the time resolution is 1 day, day-level time units are generated. Then, the WRF model and WRF-CMAQ model combine the spatial and temporal units to generate multiple spatiotemporal units and output the corresponding meteorological and air pollution data within each spatiotemporal unit.

[0055] To meet the requirements of model simulation, the following information needs to be obtained and the corresponding settings need to be completed:

[0056] (1) The information required for the operation of the WRF model includes reanalysis data for providing meteorological initial field and boundary field information, such as FNL reanalysis data released by the National Center for Environmental Prediction in the United States, which can be obtained through the center's official website or related data sharing platforms.

[0057] Before running a WRF model, conditions such as simulation area, spatial resolution, temporal resolution, number of vertical layers, physical parameterization scheme, and output frequency need to be set.

[0058] After the WRF model is run, it outputs meteorological data such as temperature, humidity, and air pressure in different spatiotemporal units. In this example, the temperature data is near-surface temperature and the humidity data is relative humidity.

[0059] (2) The information required for the CMAQ model to run includes meteorological data output by the WRF model and the pollution source emission inventory. Among them, the meteorological data output by the WRF model needs to be processed by the MCIP module and converted into a CMAQ readable format. The pollution source emission inventory can be the MEIC emission inventory, but in order to obtain higher accuracy simulation results, this study example preferably uses emission data after the MEIC emission inventory has been processed with high spatiotemporal resolution by the ISAT model; the ISAT model performs fine allocation of the MEIC emission inventory in time and space dimensions, and its output emission data has preset time resolution (such as hourly level) and spatial resolution (such as kilometer level), which can be consistent with the spatiotemporal resolution of the WRF model and the CMAQ model according to specific research needs, or set separately according to the activity characteristics of the emission sources.

[0060] Before running a CMAQ model, conditions such as simulation region, chemical mechanism, initial and boundary conditions, spatial resolution, temporal resolution, and output frequency need to be set.

[0061] After the CMAQ model is run, it outputs atmospheric pollution data such as the concentration of particulate matter (PM2.5) and its chemical components in different spatiotemporal units. The chemical components of particulate matter (PM2.5) typically include water-soluble ions (such as SO42-, NO3-, NH4-, Cl-). - Na + K + Ca 2+ Mg 2+ The components include organic carbon (OC) and elemental carbon (EC), as well as inorganic elements (such as crustal elements Al, Si, Ca, Fe, Ti, etc. and trace elements Pb, Zn, Cu, Mn, V, Ni, etc.). In actual research, appropriate indicators can be selected from the above components for subsequent analysis according to research needs. For example, crustal elements, water-soluble ions, or carbon can be selected as indicators of different types of air pollution to assess their impact on the risk of tuberculosis.

[0062] S2. Identify extreme weather events based on the meteorological data obtained in step S1, identify air pollution events based on the air pollution data obtained in step S1, and identify composite exposure events based on the meteorological data and air pollution data obtained in step S1. A composite exposure event is defined as a combination of an extreme weather event and an air pollution event that occur at a time interval of less than a preset threshold and are spatially co-located.

[0063] In this study, the main focus is on the impact of temperature on the risk of tuberculosis incidence. Therefore, the extreme weather events are primarily focused on extreme high-temperature events and extreme low-temperature events. To conduct a more refined assessment, this study further distinguishes extreme high-temperature events into daytime hot events and nighttime hot events, and extreme low-temperature events into daytime cold events and nighttime cold events. The specific identification process is as follows: (1) Extract the daily maximum and minimum temperatures within each spatial unit from the temperature data, and construct daily maximum temperature sequences and daily minimum temperature sequences respectively; (2) Set the A percentile of the daily maximum temperature sequence as the first temperature threshold, the B percentile of the daily maximum temperature sequence as the second temperature threshold, the C percentile of the daily minimum temperature sequence as the third temperature threshold, and the D percentile of the daily minimum temperature sequence as the fourth temperature threshold; (3) Make daily judgments: if the daily maximum temperature of a certain day is ≥ the first temperature threshold, then it is determined that a daytime hot event has occurred on that day; if the daily maximum temperature of a certain day is ≤ the second temperature threshold, then it is determined that a daytime cold event has occurred on that day; if the daily minimum temperature of a certain day is ≥ the third temperature threshold, then it is determined that a nighttime hot event has occurred on that day; if the daily minimum temperature of a certain day is ≤ the fourth temperature threshold, then it is determined that a nighttime cold event has occurred on that day.

[0064] The values ​​of A, B, C, and D are determined based on actual research needs. In meteorology, A and C are usually set to 90, i.e., the 90th percentile, to identify extreme high-temperature events; B and D are set to 10, i.e. the 10th percentile, to identify extreme low-temperature events.

[0065] This study primarily examines the impact of atmospheric particulate matter (PM2.5) chemical composition concentrations on the risk of tuberculosis incidence. Based on different sources and formation mechanisms of air pollution, this study categorizes air pollution events into four types: dust pollution events, secondary pollution events, combustion pollution events, and cumulative pollution events. The identification indicators for these four types are as follows: Crustal elements in PM2.5 mainly originate from soil dust and mineral dust; therefore, crustal elements are used as an indicator for dust pollution events. Water-soluble ions in PM2.5 are primarily secondary pollutants generated from gaseous precursors through atmospheric chemical reactions; therefore, water-soluble ions are used as an indicator for secondary pollution events. Carbonaceous matter in PM2.5 mainly originates from fossil fuel combustion and biomass combustion; therefore, carbonaceous matter is used as an indicator for combustion pollution events. When two or three of the above air pollution events occur simultaneously, it is identified as a cumulative pollution event.

[0066] The specific identification process for various air pollution events is as follows: (1) Extract the daily total concentration data of crustal elements, total concentration data of water-soluble ions, and total concentration data of carbonaceous matter in PM2.5 within each spatial unit from the air pollution data, and construct daily dust-type pollutant concentration sequences, daily secondary pollutant concentration sequences, and daily combustion-type pollutant concentration sequences, respectively; (2) Set the E percentile of the daily dust-type pollutant concentration sequence as the first pollutant threshold, set the F percentile of the daily secondary pollutant concentration sequence as the second pollutant threshold, and set the G percentile of the daily combustion-type pollutant concentration sequence as the third pollutant threshold. Threshold; (3) Daily judgment: If the daily concentration of floating dust pollutants on a certain day is ≥ the first pollutant threshold, then it is judged that a floating dust pollution event has occurred on that day; if the daily concentration of secondary pollutants on a certain day is ≥ the second pollutant threshold, then it is judged that a secondary pollution event has occurred on that day; if the daily concentration of combustion pollutants on a certain day is ≥ the third pollutant threshold, then it is judged that a combustion pollution event has occurred on that day; if a certain day meets two or more of the following conditions: daily floating dust pollutant concentration ≥ the first pollutant threshold, daily secondary pollutant concentration ≥ the second pollutant threshold, and daily combustion pollutant concentration ≥ the third pollutant threshold, then it is judged that a cumulative pollution event has occurred on that day.

[0067] The values ​​of E, F, and G are determined based on actual research needs. The specific values ​​may vary depending on the research area, seasonal characteristics, or research objectives. No further specific limitations are made here.

[0068] Based on the foregoing, identifying compound exposure events requires both spatial co-location and temporal proximity. Spatial co-location means that extreme weather events and air pollution events occur within the same spatial unit. Temporal proximity is determined by a preset threshold, which is set based on the time series similarity analysis results of extreme weather events and air pollution events.

[0069] Specifically, the Jaccard similarity between extreme weather events and air pollution events at different time intervals is first calculated, and the trend of its change with the time interval is analyzed. When the Jaccard similarity tends to stabilize in the short term, the time interval corresponding to the highest Jaccard similarity is taken as the preset threshold. When a certain type of extreme weather event and a certain type of air pollution event occur in the same spatial unit, and the time interval between their occurrences is less than the preset threshold, it is determined that a combined exposure event of that type has occurred.

[0070] The specific operation method for determining the preset threshold using Jaccard similarity is as follows: First, for each type of extreme weather event and air pollution event, the occurrence dates within the target time period are counted to construct an event occurrence set; then, the range of occurrence time intervals to be examined is set. For each occurrence time interval, one type of event set is used as the benchmark, and the other type of event set is time-shifted. The Jaccard similarity between the two sets after shifting is calculated, which is the ratio of the intersection to the union. This value reflects the degree of time overlap between the two types of events under this occurrence time interval; the Jaccard similarity values ​​corresponding to different occurrence time intervals are plotted as curves. As the occurrence time interval increases, the change in the Jaccard similarity value tends to be gradual. The occurrence time interval corresponding to the maximum value in this process is taken as the preset threshold.

[0071] In this study example, when determining a combined exposure event, the order of occurrence of extreme weather events and air pollution events is not considered; the criterion is solely whether the time interval between their occurrences is less than a preset threshold. Specifically, using the time unit of one type of event as a baseline, and the preset threshold as the time window width, the preset threshold duration is shifted forward from the start of that time unit and backward from its end, thus determining the candidate time interval. If another type of event occurs within the same spatial unit within this candidate time interval, then the combined exposure event of that type occurs in the time unit containing the later of the two events.

[0072] Based on the foregoing, extreme weather events include four types: daytime hot events, daytime cold events, nighttime hot events, and nighttime cold events. Air pollution events include four types: dust pollution events, secondary pollution events, combustion pollution events, and cumulative pollution events. Combining these two types yields sixteen types of composite exposure events: daytime hot-dust composite exposure events, daytime hot-secondary composite exposure events, daytime hot-combustion composite exposure events, daytime hot-cumulative composite exposure events; daytime cold-dust composite exposure events, daytime cold-secondary composite exposure events, daytime cold-combustion composite exposure events, daytime cold-cumulative composite exposure events; nighttime hot-dust composite exposure events, nighttime hot-secondary composite exposure events, nighttime hot-combustion composite exposure events, nighttime hot-cumulative composite exposure events; nighttime cold-dust composite exposure events, nighttime cold-secondary composite exposure events, nighttime cold-combustion composite exposure events, and nighttime cold-cumulative composite exposure events.

[0073] S3. Based on the principle of spatial co-location, a regression model is constructed using the number of cases in the pulmonary tuberculosis incidence data obtained in step S1 as the dependent variable and whether the composite exposure event identified in step S2 occurs as a binary explanatory variable to evaluate the impact of the composite exposure event on the risk of pulmonary tuberculosis incidence.

[0074] It should be noted that the above "based on spatial colocation" means that the combined exposure event and the incidence of pulmonary tuberculosis must be consistent in spatial units, which is the basis for matching the two. In terms of time dimension, ideally, the two should be matched within the same time unit. However, since the incidence of pulmonary tuberculosis may have a lag effect, it often takes a period of time after the occurrence of the combined exposure event for the number of cases to change. Therefore, the time matching rules can be flexibly set according to the research needs: the combined exposure event and the incidence of pulmonary tuberculosis can be the same in time unit, or there can be a certain time lag. The lag time window can be set according to the disease incubation period, the mechanism of action of the exposure factor and the research purpose. For example, it can be set as the number of cases within a certain number of days after the occurrence of the combined exposure event as the dependent variable.

[0075] The regression model used in this study is a generalized additive model of a quasi-Poisson distribution, and its analytical expression is:

[0076] logE(N) = βZ coexpo +s(RH,k1)+s(PRS,k2)+ns(time,df)+DOW+sex+age+ε;

[0077] In the formula: N is the number of cases of pulmonary tuberculosis in the incidence data, and E(N) is the expected value of N; Z coexpo Z is a binary explanatory variable indicating whether the combined exposure event occurred, where 1 indicates occurrence and 0 indicates non-occurrence; β is an explanatory variable Z. coexpoThe corresponding regression coefficients reflect the degree of influence of the combined exposure event on the incidence rate; s(RH,k1) and s(PRS,k2) are spline smoothing functions of the humidity data RH and the air pressure data PRS in the meteorological data, used to control the possible nonlinear effects of these two meteorological factors, with k1 and k2 being the corresponding degrees of freedom; ns(time,df) is the natural cubic spline function of the time trend time, used to control long-term time trends and seasonal fluctuations, with df being the degree of freedom; DOW is a dummy variable to control the weekday effect, used to eliminate the influence of differences between weekdays and weekends; sex is a binary gender variable based on the incidence data of pulmonary tuberculosis, used as a dummy variable when included in the model to control the confounding effect of gender on the incidence of pulmonary tuberculosis; age is an age classification variable based on the incidence data of pulmonary tuberculosis, and the age classification criteria and the specific number of levels included in the model can be set according to the actual research needs, for example, it can be divided into several levels such as 0-10 years, 11-20 years, and 21-30 years, and included in the model as a dummy variable; ε is the intercept term.

[0078] After the model is fitted, the relative risk (RR) = exp(β) is calculated using the regression coefficient β to quantify the risk of tuberculosis incidence from different combined exposure events. The magnitude of the RR value directly reflects the strength of the association between the combined exposure event and tuberculosis incidence: if RR > 1 and is statistically significant, it indicates that this type of combined exposure event significantly increases the risk of tuberculosis incidence.

[0079] Based on the above assessment results, this invention can be further applied to early warning and prevention of tuberculosis incidence risk and environmental risk management. Specifically, based on the relative risk (RR) values ​​corresponding to different combined exposure events, the types of combined exposure events that contribute significantly to the incidence of tuberculosis can be identified. By combining meteorological forecast data and air pollution forecast data, it is possible to predict whether such combined exposure events will occur within a certain spatiotemporal unit in the future, thereby identifying high-risk periods and high-risk areas for tuberculosis incidence. This provides a scientific basis for public health departments to conduct targeted early warning and early intervention for key populations, such as strengthening health management of key populations during periods of overlapping extreme weather and pollution processes. At the same time, by identifying the changing characteristics of tuberculosis incidence risk under combined exposure conditions, this invention also provides information support for environmental management departments to assess combined environmental risks and formulate differentiated risk management strategies.

[0080] The following study uses Tianjin as the target area to assess the risk of tuberculosis incidence under combined exposure to extreme weather and air pollution.

[0081] S1. Obtain meteorological data, air pollution data, and tuberculosis incidence data for Tianjin from 2015 to 2024.

[0082] (1) Meteorological data include temperature data, humidity data and air pressure data. The data were simulated using the WRF model version 4.0 based on FNL reanalysis data, with a spatial resolution of 9km×9km and a temporal resolution of 1h.

[0083] (2) The atmospheric pollution data includes the chemical component concentration data of particulate matter PM2.5, specifically including the total concentration data of crustal elements, the total concentration data of water-soluble ions and the total concentration data of carbonaceous matter in particulate matter PM2.5. The data are obtained by coupling the WRF model version 4.0 and the CMAQ model version 5.3.1, and are based on FNL reanalysis data and pollution source emission inventory. The pollution source emission inventory is the emission data after the MEIC emission inventory has been processed with high spatiotemporal resolution by the ISAT model. The chemical reaction mechanism used in the CMAQ model is the CB6r3_AE7_AQ mechanism, with a spatial resolution of 9km×9km and a temporal resolution of 1h.

[0084] (3) Tuberculosis incidence data include the daily incidence rate, as well as the onset time, spatial location, gender, age, etc. of the cases, which are obtained through disease control departments and medical institutions.

[0085] S2. Identify extreme weather events based on the meteorological data obtained in step S1, identify air pollution events based on the air pollution data obtained in step S1, and identify composite exposure events based on the meteorological data and air pollution data obtained in step S1. A composite exposure event is defined as a combination of an extreme weather event and an air pollution event that occur at a time interval of less than a preset threshold and are spatially co-located.

[0086] (1) Extract the daily maximum and minimum temperatures within each spatial unit from the temperature data, and construct the daily maximum temperature sequence and the daily minimum temperature sequence respectively; (2) Set the 90th percentile of the daily maximum temperature sequence as the first temperature threshold, set the 10th percentile of the daily maximum temperature sequence as the second temperature threshold, set the 90th percentile of the daily minimum temperature sequence as the third temperature threshold, and set the 10th percentile of the daily minimum temperature sequence as the fourth temperature threshold; (3) Make daily judgments: if the daily maximum temperature of a certain day is ≥ the first temperature threshold, then it is judged that a daytime hot event has occurred on that day; if the daily maximum temperature of a certain day is ≤ the second temperature threshold, then it is judged that a daytime cold event has occurred on that day; if the daily minimum temperature of a certain day is ≥ the third temperature threshold, then it is judged that a nighttime hot event has occurred on that day; if the daily minimum temperature of a certain day is ≤ the fourth temperature threshold, then it is judged that a nighttime cold event has occurred on that day.

[0087] (2) Extract the daily total concentration data of crustal elements, total concentration data of water-soluble ions, and total concentration data of carbonaceous matter in PM2.5 within each spatial unit from the atmospheric pollution data, and construct daily dust-type pollutant concentration sequences, daily secondary pollutant concentration sequences, and daily combustion-type pollutant concentration sequences, respectively; (2) Set the 80th percentile of the daily dust-type pollutant concentration sequence as the first pollutant threshold, set the 75th percentile of the daily secondary pollutant concentration sequence as the second pollutant threshold, and set the 75th percentile of the daily combustion-type pollutant concentration sequence as the third pollutant threshold; (3) Step by step Daily assessment: If the daily concentration of particulate matter pollutants on a certain day is greater than or equal to the first pollutant threshold, then a particulate matter pollution event is determined to have occurred on that day; if the daily concentration of secondary pollutants on a certain day is greater than or equal to the second pollutant threshold, then a secondary pollution event is determined to have occurred on that day; if the daily concentration of combustion pollutants on a certain day is greater than or equal to the third pollutant threshold, then a combustion pollution event is determined to have occurred on that day; if a certain day meets two or more of the following conditions: daily particulate matter pollutant concentration greater than or equal to the first pollutant threshold, daily secondary pollutant concentration greater than or equal to the second pollutant threshold, or daily combustion pollutant concentration greater than or equal to the third pollutant threshold, then a cumulative pollution event is determined to have occurred on that day.

[0088] In calculating the percentile thresholds for each of the above sequences, a sliding window method was used: for each day, data from that day and the seven days before and after it (a total of 15 days) were used to construct a sliding sequence, and the percentile thresholds for that day were calculated based on this sliding sequence. That is, the first to fourth temperature thresholds and the first to third pollutant thresholds are all daily dynamic thresholds. The daily thresholds are calculated based on the data from the same period of that day and the seven days before and after it, to ensure that event identification can reflect the degree of anomaly of the day relative to the climate and pollution background of the same period.

[0089] (3) Based on the extreme weather events and air pollution events identified above, further identify compound exposure events.

[0090] First, the Jaccard similarity between extreme weather events and air pollution events was calculated for different time intervals. The calculations revealed that the Jaccard similarity was highest when the time interval was 2 days; therefore, 2 days was used as the preset threshold.

[0091] Based on the principle of spatial co-location, for the same spatial unit, if the time interval between a certain type of extreme weather event and a certain type of air pollution event is ≤2 days, then the time unit in which the latter event occurs is taken as the occurrence time of the composite exposure event, and it is determined that a composite exposure event of the corresponding combination type has occurred within that time unit.

[0092] Based on the aforementioned event types, composite exposure events include daytime heat-dust composite exposure events, daytime heat-secondary composite exposure events, daytime heat-combustion composite exposure events, daytime heat-cumulative composite exposure events, daytime cold-dust composite exposure events, daytime cold-secondary composite exposure events, daytime cold-combustion composite exposure events, daytime cold-cumulative composite exposure events, nighttime heat-dust composite exposure events, nighttime heat-secondary composite exposure events, nighttime heat-combustion composite exposure events, nighttime heat-cumulative composite exposure events, nighttime cold-dust composite exposure events, nighttime cold-secondary composite exposure events, nighttime cold-combustion composite exposure events, and nighttime cold-cumulative composite exposure events.

[0093] The composite exposure events in Tianjin from 2015 to 2024 were identified and statistically analyzed using the above method. The frequency of occurrence of each type of composite exposure event in each spatial unit was obtained. Figure 1 The frequency spatial distribution of four typical combined exposure event types is shown in the figure, with the color intensity representing the frequency of occurrence. From Figure 1 It can be seen that the frequency of different types of combined exposure events varies significantly in different areas of Tianjin. This reflects the spatial distribution characteristics of combined exposure events and can provide spatial visualization basis for public health departments to carry out differentiated early warning and prevention and control strategies for different areas.

[0094] S3. Based on the principle of spatial colocation, using the daily incidence rate in the pulmonary tuberculosis incidence data obtained in step S1 as the dependent variable and whether the combined exposure event identified in step S2 occurs as the binary explanatory variable, a generalized additive model with a quasi-Poisson distribution is constructed to assess the impact of the combined exposure event on the risk of pulmonary tuberculosis incidence.

[0095] The analytical expression of the generalized additive model corresponding to various types of composite exposure events is uniformly represented by the logarithmic link function, that is:

[0096] logE(N) = βZ coexpo +s(RH,k1)+s(PRS,k2)+ns(time,df)+DOW+sex+age+ε;

[0097] In the formula: N is the daily incidence rate of pulmonary tuberculosis, and E(N) is the expected value of N; Z coexpo Z is a binary explanatory variable indicating whether the combined exposure event occurred, where 1 indicates occurrence and 0 indicates non-occurrence; β is an explanatory variable Z. coexpoThe corresponding regression coefficients reflect the degree of influence of the combined exposure event on the incidence rate; s(RH,k1) and s(PRS,k2) are spline smoothing functions of the humidity data RH and the air pressure data PRS in the meteorological data, used to control the possible nonlinear effects of these two meteorological factors, with k1 and k2 being the corresponding degrees of freedom, each set to 3; ns(time,df) is the natural cubic spline function of the time trend time, used to control long-term time trends and seasonal fluctuations, with df being the degrees of freedom, set to 7 per year; DOW is a dummy variable to control the weekday effect; sex is a binary gender variable based on the incidence data of pulmonary tuberculosis, included in the model as a dummy variable to control the confounding effect of gender on the incidence of pulmonary tuberculosis; age is an age classification variable based on the incidence data of pulmonary tuberculosis, and the age classification criteria and the specific number of levels included in the model can be set according to actual research needs, for example, it can be divided into several levels such as 0-10 years, 11-20 years, and 21-30 years, and included in the model as a dummy variable; ε is the intercept term.

[0098] After the model is fitted, the regression coefficients β and their standard errors SE corresponding to each type of combined exposure event are output. The standard error SE is obtained from the covariance matrix of the parameter estimates and is used to measure the estimation accuracy of the regression coefficient β; a smaller SE indicates a more stable estimation of β.

[0099] Based on the regression coefficient β and its standard error SE, two core statistical indicators can be further calculated to assess the impact of the combined exposure event: First, the relative risk (RR), calculated as RR = exp(β), quantifies the impact of the combined exposure event on the risk of tuberculosis. Based on this, the 95% confidence interval for RR is calculated as exp(β ± 1.96 × SE), which reflects the range within which the true RR value might fall at a 95% confidence level. Second, a statistical significance test is performed using the t-statistic, calculated as t = β / SE, and the corresponding p-value is calculated using the t-distribution. When the p-value is less than 0.05, the association between this type of combined exposure event and the risk of tuberculosis is considered statistically significant.

[0100] Table 1 shows the RR values, 95% confidence intervals, and P values ​​for the sixteen types of combined exposure events in this study.

[0101]

[0102] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution, characterized in that, Including the following steps: Acquire meteorological data, air pollution data, and tuberculosis incidence data for the target area within the target time period; Extreme weather events are identified based on the meteorological data, air pollution events are identified based on the air pollution data, and composite exposure events are identified based on the meteorological data and the air pollution data. The composite exposure event is a combination of the extreme weather event and the air pollution event that occur at a time interval of less than a preset threshold and are spatially co-located. Based on the spatial colocation, a regression model is constructed using the number of cases in the pulmonary tuberculosis incidence data as the dependent variable and whether the combined exposure event occurs as a binary explanatory variable to assess the impact of the combined exposure event on the risk of pulmonary tuberculosis incidence.

2. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 1, characterized in that, The preset threshold is set based on the time series similarity analysis results of the extreme weather event and the air pollution event.

3. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 2, characterized in that, The time series similarity analysis includes: calculating the Jaccard similarity between the extreme weather event and the air pollution event under different occurrence time intervals, and using the occurrence time interval corresponding to the highest Jaccard similarity as the preset threshold.

4. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 1, characterized in that, The term "spatial co-location" refers to the extreme weather event and the air pollution event occurring within the same spatial unit, which is obtained by dividing the target area into grids.

5. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 4, characterized in that, The assessment methods for the extreme weather events include: Based on the meteorological data, construct the daily maximum temperature sequence and the daily minimum temperature sequence within the spatial unit; If the daily maximum temperature is greater than or equal to the first temperature threshold, it is a daytime heat event, where the first temperature threshold is the A percentile of the daily maximum temperature sequence; If the daily maximum temperature is less than or equal to the second temperature threshold, it is a daytime cold event, where the second temperature threshold is the B percentile of the daily maximum temperature sequence. If the daily minimum temperature is greater than or equal to the third temperature threshold, it is a nighttime heat event, where the third temperature threshold is the C percentile of the daily minimum temperature sequence; If the daily minimum temperature is less than or equal to the fourth temperature threshold, it is considered a cold night event, where the fourth temperature threshold is the D percentile of the daily minimum temperature sequence.

6. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 5, characterized in that, The meteorological data includes temperature, air pressure, and humidity data obtained by simulation using the WRF model based on FNL reanalysis data.

7. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 4, characterized in that, The assessment methods for the aforementioned air pollution events include: Based on the aforementioned air pollution data, daily dust-type pollutant concentration sequences, daily secondary pollutant concentration sequences, and daily combustion pollutant concentration sequences are constructed within the aforementioned spatial unit. If the daily concentration of particulate matter is greater than or equal to the first pollutant threshold, then it is a particulate matter pollution event, where the first pollutant threshold is the E percentile of the daily particulate matter concentration sequence. If the daily concentration of secondary pollutants is greater than or equal to the second pollutant threshold, then it is a secondary pollution event, where the second pollutant threshold is the F percentile of the daily secondary pollutant concentration sequence. If the daily concentration of combustion pollutants is greater than or equal to the third pollutant threshold, then it is a combustion pollution event, where the third pollutant threshold is the G percentile of the daily combustion pollutant concentration sequence. If two or more of the following conditions are met: daily concentration of particulate matter ≥ the first pollutant threshold, daily concentration of secondary pollutants ≥ the second pollutant threshold, or daily concentration of combustion pollutants ≥ the third pollutant threshold, then it is considered a cumulative pollution event.

8. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 7, characterized in that, The air pollution data includes particulate matter (PM2.5) obtained by simulation using the WRF-CMAQ model based on FNL reanalysis data and pollution source emission inventories. 2.5 and its chemical component concentration data; the particulate matter PM 2.5 The chemical composition includes crustal elements for identifying dust-type pollution events, ions for identifying secondary pollution events, and carbonaceous matter for identifying combustion-type pollution events.

9. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to any one of claims 1-8, characterized in that, The regression model is a generalized additive model of a quasi-Poisson distribution, and its analytical expression is: logE(N)=βZ coexpo +s(RH,k1)+s(PRS,k2)+ns(time,df)+DOW+sex+age+ε; In the formula: E(N) is the expected value of the number of cases N in the pulmonary tuberculosis incidence data; Z coexpo It is a binary explanatory variable for whether the composite exposure event occurred; β is the explanatory variable Z coexpo The corresponding regression coefficients; s(RH,k1) and s(PRS,k2) are spline smoothing functions for the humidity data RH and the air pressure data PRS in the meteorological data, where k1 and k2 represent the degrees of freedom. ns(time,df) is the natural cubic spline function of the time trend time, where df represents the degrees of freedom; DOW is a dummy variable used to control for the weekday effect; sex is a binary gender variable based on the tuberculosis incidence data; age is an age grading variable based on the tuberculosis incidence data; ε is the intercept term.

10. The method for assessing the risk of pulmonary tuberculosis incidence under combined exposure to extreme weather and air pollution according to claim 9, characterized in that, The impact of the combined exposure event on the risk of developing pulmonary tuberculosis was assessed based on the relative risk (RR), which is expressed as RR = exp(β).

Images