{"paper_id":"64f63ffc-b166-4754-a2df-59153969bb1c","body_text":"Mapping the global distribution of and environmental suitability for scrub typhus | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Mapping the global distribution of and environmental suitability for scrub typhus Richard Maude, Qian Wang, Tian Ma, Fangyu Ding, Ivo Elliott, Canjun Zheng, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6251403/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Scrub typhus, an acute febrile illness caused by Orientia tsutsugamushi , has emerged as a significant public health concern, expanding beyond its traditional endemic region, the \"tsutsugamushi triangle\" in the Asia-Pacific. Despite its increasingly global distribution, comprehensive spatial assessments of scrub typhus risk remain sparse. An exhaustive assembly of 56,093 unique human scrub typhus occurrence records worldwide was undertaken from published literature and national surveillance datasets. Covering 27 countries/regions, these records were combined with 28 climatic, geographic, and socio-economic covariates environmental covariates using an ensemble machine learning modelling approach, capturing possible nonlinear effects and complex interactions, to map the probability of occurrence at 5×5 km resolution globally. This approach involved stacking of three sub-models (generalized additive models, boosted regression trees and random forest). Environmental suitability for scrub typhus was found to be highest in moderate to tropical climates, notably extending beyond the classic \"tsutsugamushi triangle\" into large sections of Central and South America, Central and West Africa. Approximately 2.5 billion people (95% CI: 2.43–2.69 billion) are estimated to be currently living in environmentally suitable areas within countries or regions where human cases of scrub typhus have already been confirmed. This number increases to 4.4 billion people (95% CI: 3.86–4.90 billion) if countries without confirmed cases are included. This data assembly and modelled environmental suitability surface provide novel insights into the potential public health impact of scrub typhus. This may serve as a catalyst for broader discussions regarding the neglected global impact of this disease, the need to improve public awareness, drug, and vector control methods, and lead to further burden assessment. The study highlights key data gaps, particularly in regions with limited surveillance and accessibility of healthcare facilities, and emphasizes the need for future research in the context of ongoing climate and environmental changes, which may further alter the global distribution of scrub typhus. Health sciences/Diseases/Infectious diseases/Bacterial infection Health sciences/Risk factors Health sciences/Medical research/Epidemiology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Scrub typhus, also known as tsutsugamushi disease, is an acute febrile illness and a neglected vector-borne infectious disease that poses a serious public health threat globally 1 . It is mainly caused by the bacterium Orientia tsutsugamushi and is transmitted to humans through the bite of the larval stage of trombiculid mites, “chiggers” 2 . The disease presents with symptoms ranging from fever, headache, and rash to more severe complications such as pneumonitis, meningoencephalitis, and multi-organ failure 3 , making it a critical public health concern. Originally, the endemic area for scrub typhus was considered to be restricted to the geographic “tsutsugamushi triangle” which extends from the far east of Russia in the north, to northern Australia in the south, and Afghanistan and Pakistan in the west. 4 Many countries in this region have suffered from scrub typhus for a long time, particularly China, Japan, Korea, Thailand, and India 5-8 . But there is mounting evidence to suggest that the range of scrub typhus expands beyond these traditional boundaries to the Middle East, South America, and Africa 9-11 . Recently, two new Orientia species causing scrub typhus were isolated in unexpected areas: Candidatus Orientia chuto in Dubai 12 and Candidatus Orientia chiloensisin Chile 13 , which not only indicates a geographic expansion, but may also imply a lack of awareness and attention to the disease outside of the triangle. Detailed spatial information on the global geographical extent of scrub typhus and areas that are environmentally suitable for transmission to humans remains sparse. Previous studies have primarily assessed scrub typhus at national level 14,15 , with risk maps generated only for a few countries or subnational units 16 , 17,18 . Given recent improvement in empirical data as well as advances in disease modelling approaches, a comprehensive assessment of scrub typhus records and the creation of high-resolution environmental suitability maps on a global scale was carried out. Currently, spatial analyses of scrub typhus are few in number compared to some other neglected tropical diseases (NTDs). Several studies have attempted to identify the chief determinants of scrub typhus geographic distribution patterns 19-25 . Strong correlations have been found in China, Thailand, South Korea, Japan and Nepal between scrub typhus occurrence and suitable environments, including meteorological factors, geographic factors, socioeconomic variables, and biological factors. Temperature, precipitation, and humidity, as well as the El Niño/Southern Oscillation (ENSO) and Multivariate ENSO Index, have been found to be associated with scrub typhus in China 26-28 . Geographic indicators have included elevation, Normalized Difference Vegetation Index (NDVI), cropland, forest, and grassland landcovers. Deforestation exhibited positive associations with scrub typhus incidence in South Korea. One study in Thailand showed habitat complexity and fragmentation to be significant predictors for scrub typhus cases 29,30 . The numbers of patients with scrub typhus infection were found to be higher in villages with higher levels of surface flooding and vegetation in a study in Laos 22 . Socioeconomic features have also been associated with spatial patterns of scrub typhus risk, such as a negative association between travel time to cities and reported cases in China 31 . Biological factors, such as rodent density and Orientia infection rate in mites/rodent also have been examined to have association with scrub typhus 27,32,33 . Knowledge of the geographical distribution of potential scrub typhus infection risk is critical for identifying and prioritizing regions for preventive and targeted control measures and is essential for estimating the global clinical burden of scrub typhus, which is currently unknown. Therefore, a comprehensive database of timepoints and locations where scrub typhus has been reported was developed. Then these data were coupled with a comprehensive set of environmental covariates to generate high-resolution risk maps (5km x 5km), representing global environmental suitability for scrub typhus occurrence. Three single machine learning models and two stacked ensemble methods were applied to generate these risk maps with 95% uncertainty intervals. By developing a global environmental suitability map, the aim was to provide insights into the potential spatial distribution of scrub typhus risk and identify potential hotspots for further investigation, targeted surveillance and control measures. The results of this study will contribute to the development of evidence-based strategies to mitigate the impact of this neglected tropical disease. Methods Overview/Summary A comprehensive global database of geo-located scrub typhus occurrence records was compiled, integrating data from published literature and national surveillance systems, and standardized to annual intervals. Absence data were rigorously verified. Environmental and socio-economic covariates were processed to a unified 5 x 5 km resolution. An ensemble machine learning approach, including generalized additive models, boosted regression trees, and random forests, was conducted combining multiple models to predict the probability of occurrence (with uncertainty) as a measure of environmental suitability. The population living in suitable area was estimated by spatially linking the predicted environmental suitability with high-resolution human population data, enabling an assessment of populations potentially exposed to scrub typhus. The GATHER(Guidelines for Accurate and Transparent Health Estimates Reporting) checklist was followed to report the description of input data and estimate method 34 (Supplementary Table S1). The occurrence database A database of 67,905 geo-located occurrence records was initially compiled for the period 1944 to 2024, drawing on a diverse array of sources including published literature, online repositories, and national surveillance systems. After thorough standardization, quality control, and deduplication, the final dataset comprised 56,093 unique year-location occurrences, serving as the basis for modelling. A systematic literature search was conducted to capture all relevant occurrences, irrespective of language, publication date, or geographical focus. The processes of searching, screening, data extraction, cleaning, and geolocation have been thoroughly described in our earlier published work 35 . The occurrence database was first created from published literature, case reports/series and grey literatures and was last updated in May 2024. This extensive review yielded data from 829 published references, resulting in the extraction of 7,165 unique year-location occurrence datapoints, which were subsequently subjected to temporal standardization and spatial processing. As scrub typhus remains a notifiable disease in several Asia countries/regions, data was also sourced from national surveillance systems. We retrieved official records from the online public databases of Japan, South Korea, and Taiwan, and secured access to unpublished national data from Mainland China and Thailand through direct correspondence with relevant health authorities. Detailed methodologies pertaining to data acquisition and processing from these sources are provided in the Supplementary Information section 1.1. All records were standardized annually by location, meaning that repeated records from the same location within a single year were consolidated into a singular occurrence and underwent rigorous quality control, as outlined in Supplementary Information section 1.2. The final occurrence database contained 56,093 unique year-location occurrences, which represent a unique location where one or more cases occurred within one year. A map of the final set of occurrence locations utilized for modelling the contemporary distribution of risk for scrub typhus is provided in Supplementary Fig. S1(a), with the number of occurrences per year globally and by country/region is shown in Supplementary Fig. S1(b). Background location database Our methods require both presence and absence data to define areas of disease absence and potentially unsuitable environmental conditions at unsampled locations. To ensure the absence data is as accurate as possible, it was sourced from both published literature and national surveillance systems. The negative records were extracted from the published sources and cross checked their geolocations with other published literature and national surveillance data by applying dual approach tailored to the nature of the original data. For point-based data, a 30 km radius was applied to ensure no positive or reported cases were present within the surrounding area. For area-based data, the verification was conducted within the same administrative polygons to confirm the absence of reported cases. Additionally, a significant proportion of absence data came from national official reports. Multiple years of data were examined for the smallest administrative divisions, retaining units that had never reported any cases, whether suspected probable, or confirmed. These locations were also cross-checked with literature records to ensure no positive cases were documented. The final absence data comprised 122 published literature records from 13 countries/regions, 400,829 reported absence year-county records from the China CDC, 476 reported year-district records from the Taiwan CDC, and 384 reported year-district records from the Thailand Division of Vector Borne Diseases (DVBD). To balance the presence-absence ratio for modelling, a subset of absence records was randomly selected from this database, ensuring a proportionate distribution relative to the 56,093 initial presence records. Covariates Given that humans are typically incidental hosts and that the bacterium causing scrub typhus is primarily maintained in, and transmitted by, chigger mites within natural cycles, the activities and distribution of these vectors, the presence, movements and population fluctuations of small mammal hosts—particularly rodents and human behaviours, are inherently linked to the occurrence of the disease. These factors are strongly influenced by climatic and other environmental conditions. To identify indicators that have been shown previously to have significant associations with scrub typhus, a systematic review was conducted to extract existing evidence 19 . This systematic review, which analysed data from 68 articles published between 1978 and 2024 across seven countries/regions, identified 68 significant environmental risk factors associated with scrub typhus with temperature, precipitation, humidity, sunshine duration, elevation, the normalized difference vegetation index (NDVI), the proportion of cropland, population density, and urban status as the top-ten indicators most mentioned. Based on those findings and the availability of high-resolution spatial and temporal data, 28 covariates were selected for further analysis encompassing meteorological, geographic, and socioeconomic factors. Meteorological factors included minimum temperature, maximum temperature, accumulated precipitation, relative humidity, surface air pressure, and wind speed. Geographical predictors included the proportions of 17 land cover classes, NDVI, Enhanced Vegetation Index (EVI), and elevation. Socio-economic covariates included travel time to major cities (urban accessibility) and population density. Detailed descriptions and sources of these covariates can be found in the Supplementary Information section 2 and Table S4. All raster data were aggregated and resampled to a unified spatial resolution of 5 km x 5 km and a yearly temporal resolution (Figure S2). Spatial processin g Various data formats were converted and integrated into raster format Where necessary, input data sources were re-projected using a standardized equirectangular Plate Carrée projection under the World Geodetic System 1984 (WGS 84) coordinate system. Input grids with spatial resolutions different from 5 km x 5 km were either aggregated or disaggregated to this target resolution using bilinear or nearest-neighbour interpolation techniques. Temporal processing All data were standardized to annual intervals. Temporal interpolation was not required for most covariates. Elevation and urban accessibility were treated as static covariates, with elevation considered constant over time and urban accessibility data was available only for the year 2015. Data merging For polygon-based occurrence data from surveillance systems, the mean covariate values within each polygon were calculated. For literature-reported occurrences recorded as hospital locations, the mean covariate value within a 30 km buffer zone around the hospital was calculated. This buffer distance was determined by analysing the distance between reported hospitals and onset locations from national surveillance systems across five countries/regions. For point location data, the covariate values at the exact location were used. All these processes were carried out using ArcGIS (10.8.2) and Python (3.9.13), utilizing the following packages: pandas, geopandas, rasterio, shapely, rasterstats, pyproj, and pygrib. Model An ensemble machine learning approach 36 was implemented to predict the environmental suitability for scrub typhus based on environmental conditions sampled at each site from the covariate suite, capturing potential nonlinear effects and complex interactions among covariates. After conducting multiple rounds of experimentation, Generalized Additive Models (GAM) 37 , Boosted Regression Trees (BRT) 38 , and Random Forest (RF) 39,40 were selected due to their superior performance as the three primary child models to fit the occurrence data with the covariates as predictors, as measured by the Area Under the Curve (AUC), compared to other models such as Lasso regression 41 and Generalized Linear Models (GLM) 42 . Rigorous and comprehensive model selection and parameter tuning were performed on each model to compare among a number of alternative models and enhance the performance. For GAM, covariates exhibiting high concurvity, a condition that generalizes co-linearity and can complicate model interpretation, were subsequently removed. To further enhance model robustness and mitigate the risk of overfitting, a backward stepwise procedure was employed for covariate selection. For BRT and RF, a balanced dataset approach was adopted to ensure equal representation of both presence and absence samples. This strategy was particularly important for preventing bias in the model’s predictions when dealing with imbalanced datasets. Additionally, extensive parameter tuning was conducted for each of these machine-learning methods, involving a thorough search for optimal hyperparameters (n.trees, shrinkage, interaction.depth, bag.fraction, n.minobsinnode, mtry) to maximize model performance and ensure generalization across different scenarios. All these models were carried out using R, utilizing the following packages: sf, mgcv, gbm, caret, randomForest. parallel, doParallel. The meticulous selection processes are detailed in the Supplementary Information, section 3. 70% of the whole dataset was used to train these sub-models and 30% of the data was used for out of sample validation. Each sub-model was fitted using fivefold cross-validation to avoid overfitting, and the out-of-sample predictions from the five folds were compiled into a single set of predictions that were used to fit two stacking methods: constrained weighted mean (CWM) and weighted mean based on R-square weighted mean (RWM) 43-46 . The CWM is a stacking ensemble technique that assigns weights to base models within predefined constraints to ensure stability and interpretability, balancing model complexity and predictive accuracy. The RWM assigns weights proportional to each model's R-squared value, emphasizing explanatory power. Both methods optimize the contribution of individual models, enhancing the overall performance and robustness of the ensemble. In addition, each sub-model was also fitted to the full dataset to generate a complete set of in-sample predictions that were subsequently used when generating predictions from stacked ensemble models. The mean area under the receiver operating characteristic curve (AUC) statistics among training data and validation data were used to compare the goodness of fit of those three sub-models and two stacking methods. The predicted environmental suitability was on a scale from 0 to 1, with a final prediction map generated using the best-performing model. For the uncertainty analysis, the Wilson Score interval method was employed to quantify uncertainty and calculate the 95% confidence interval 47 . To rigorously assess the robustness of the models, three strategies were implemented: (1) Bootstrap Consistency Check: a bootstrap approach 48 , where 100 distinct training datasets were generated, machine learning models executed iteratively across these datasets, and the variance meticulously analysed among the 100 predictive outputs and with the final prediction; (2) No-Time-Redundancy Prediction: integration of temporal information by ensuring that each geographic location, irrespective of its temporal reporting frequency, was uniquely represented in the model, thereby removing temporal redundancy. The adjusted dataset was then fitted using the models to generate the no-time-redundancy prediction result; and (3) Absence-Adjusted Prediction: modification of the presence-to-absence ratio, testing the model's resilience under different conditions with ratios of 1:1, 1:2, and 1:5 49 . These ratios reflect different scenarios of class imbalance, with 1:1 representing equal numbers of presence and absence data points, while 1:2 and 1:5 progressively increase the proportion of absence data to presence data. The rationale behind testing these ratios was to evaluate the model's stability and predictive power under varying degrees of absence data, as absence data often outnumbers presence data in ecological and disease models 50 . The model was refitted for each adjusted ratio to generate the absence-adjusted prediction result and assess predictive stability. The stability and robustness of the predictive outcomes were further evaluated using standard deviation metrics and corresponding maps, facilitating a comprehensive comparison of prediction variations across different scenarios. Areas were classified as \"at-risk\" based on a threshold suitability of 0.5 or greater applied to continuous suitability map values. Specifically, any pixel with a predicted suitability probability value greater than 0.5 was designated as belonging to an at-risk area. This same threshold was subsequently used to calculate the population within the delineated at-risk areas. Additionally, the areas and populations at risk were further refined by calculating confidence intervals based on the 95% confidence intervals (CI) of the environmental suitability results, providing a more robust estimate of the uncertainty associated with these predictions. Results A total of 78,532 georeferenced occurrence and absence year-locations were used to fit the models and 33,656 year-locations to validate them, of which 56,093 were occurrence records and an equal number of background points. Based on the AUC from various test methods (Table S6), the RF method was chosen for the final prediction, and subsequent testing demonstrated good robustness of the final model (Table S8 and Figure S7). Relative importance of covariates Based on the normalized relative importance of various covariates as determined by the model using the Mean Decrease Accuracy metric (Figure S5), the maximum temperature, minimum temperature, and elevation were the most influential variables. Urban and built-up areas, along with croplands, were the most significant land use types. Overall, the model exhibited a balanced reliance on a diverse set of factors rather than being dominated by one or two key variables, suggesting that multiple covariates contribute significantly to the model's predictive accuracy. Environmental Suitability Widespread environmental suitability for scrub typhus was predicted across tropical and subtropical regions (Fig. 1). The areas with highest suitability were in Southeast Asia, South Asia, northern Australia and parts of Central Africa. Areas of high suitability (probability > 0.5) were also observed across Central and South America, as well as parts of the Caribbean and Northern Australia. Smaller but notable areas of suitability were present in the Middle East, Southern Africa, and portions of East Asia. The traditional understanding of the tsutsugamushi triangle—the endemic region for scrub typhus—encompasses Southeast Asia, parts of East Asia, northern Australia and the Pacific Islands with a triangular boundary. However, the predicted environmental suitability indicates a broader potential distribution, with significant overlaps and expansions beyond the traditional tsutsugamushi triangle. This suggests that suitable environments for Orientia spp. are not confined to this traditional area but also extend beyond it into parts of South Asia, Central and Southern Africa, and South America. This suggests that the risk zones for scrub typhus could be more widespread than previously recognized, underscoring the need for enhanced global surveillance to accurately determine the true extent of the disease’s distribution. Reported incidence in high and low suitability areas Reported incidence was compared between areas of high and low environmental suitability in Mainland China, South Korea, and Japan. Thailand and Taiwan were excluded from this analysis due to uniformly high population-weighted environmental suitability scores (probability > 0.5) across all first-level administrative divisions. The results indicate that regions with environmental suitability probabilities greater than 0.5 (red) consistently exhibited higher mean annual incidences (5.0 per 100,000 population) compared to areas with scores below 0.5 (blue) (0.6 per 100,000 population) (Fig. 2A). This trend was particularly pronounced in Mainland China. A positive relationship was observed, with an increase in environmental suitability associated with a gradual rise in reported incidence as determined by a GLM passion regression analysis conducted across all first-level administrative divisions (Fig. 2B). The shaded area represents the 95% confidence interval, suggesting that while there is some variability in the data, the trend remains statistically significant, especially in higher suitability regions. Uncertainty The absolute uncertainties in the environmental suitability estimates were inherently influenced by the spatial distribution and density of occurrence data, with the highest uncertainties observed in regions with limited data or highly heterogeneous environments, notably in large areas of American, Central and Southern South American, Central Australia, plateaus, rift valleys and deserts in Africa and desert region in the Middle East (Fig. 2 and Figure S6). Furthermore, an analysis of the ratio of the mean to the width of the confidence interval highlighted the primary contributors to relative uncertainty. These were predominantly countries with sparse occurrence points and inconsistent evidence regarding the presence of scrub typhus, such as in remote or data-poor regions. Population living in environmental suitable areas Further analysis revealed significant variations in environmentally suitable areas and populations living in suitable areas across different WHO regions (Table 1). The full results for 215 countries/regions can be found in Supplementary Information, section 4.3. It was estimated that 4.4 billion people (95% CI: 3.86–4.90 billion), about 54% of the global population, live in areas with high environmental suitability for scrub typhus worldwide, while in countries/regions where scrub typhus has been confirmed in humans, 2.5 billion (95% CI: 2.43–2.69 billion) people live in environmentally suitable areas. South-East Asia, notably India and Indonesia, has extensive at-risk areas with substantial populations exposed. In the Western Pacific, China, Vietnam and Philippines were major regions of concern, while Brazil dominated in the Americas with the largest area at risk. Some African countries such as Nigeria and Ethiopia also had considerable populations living in high-suitability zones, underscoring the widespread nature of potential exposure across these diverse regions. Most countries in the Americas and Africa have never reported any confirmed human scrub typhus cases (Fig. 4A), but our analysis indicates that large areas in Brazil, the United States, Mexico, Nigeria, Ethiopia, Egypt, DR Congo and Sudan are environmentally suitable for scrub typhus. Table 1 Predicted high environmental suitability areas and top 20 highest populations living in environmentally suitable areas by WHO region. WHO region Country Area (km 2 ) Population in millions (uncertainty) Percentage of total population (uncertainty) Confirmed human scrub typhus South-East Asia India 3,231,500 1166.32 (1160.26-1169.6) 98.7% (98.2%-98.9%) Y Indonesia 1,843,375 223.89 (222.02-224.29) 97.9% (97.0%-98.0%) Y Bangladesh 147,275 139.46 (139.19-139.46) 99.1% (98.9%-99.1%) Y Thailand 531,300 57.47 (57.29-57.47) 98.9% (98.6%-98.9%) Y Myanmar (Burma) 705,525 46.14 (45.66-46.15) 97.5% (96.5%-97.6%) Y Western Pacific China 1,330,900 349.26 (301.13-427.75) 28.7% (24.7%-35.1%) Y Vietnam 338,050 80.33 (79.34-80.38) 99.2% (98.0%-99.2%) Y Philippines 282,150 75.93 (75.75-75.93) 94.1% (93.9%-94.1%) Y South Korea 109,625 40.87 (38.78-41.2) 96.4% (91.5%-97.2%) Y Americas Brazil 8,466,600 175.88 (151.74-177.57) 96.8% (83.6%-97.8%) N United States 1,659,100 150.06 (82.07-210.98) 53.5% (29.2%-75.2%) N Mexico 1,540,975 89.77 (63.29-104.9) 79.5% (56.1%-92.9%) N Africa Nigeria 924,375 174.15 (173.68-174.15) 99.1% (98.8%-99.1%) N Ethiopia 1,091,200 84.51 (64.29-90.98) 88.2% (67.1%-94.9%) N Egypt 527,375 83.61 (53.82-86.54) 96.3% (62.0%-99.7%) N DR Congo 2,254,100 76.58 (68.53-76.8) 99.6% (89.1%-99.9%) N Tanzania 931,150 51.77 (46.08-52.25) 97.7% (87.0%-98.6%) N Sudan 2,558,600 51.2 (47.94-51.23) 99.9% (93.5%-99.9%) N Kenya 579,550 42.19 (34.29-42.68) 97.8% (79.5%-98.9%) Y Eastern Mediterranean Pakistan 745,975 166.44 (148.3-169.2) 93.8% (83.6%-95.4%) Y A list of priority countries was developed where targeted research and enhanced surveillance are most urgently needed to determine if scrub typhus transmission is occurring (Fig. 4A; Supplementary Table S10). With proper surveillance in place, targeted prevention campaigns and treatment guidelines could be implemented to reveal and reduce potential burden associated with scrub typhus. The priority list includes areas where scrub typhus is likely present but underreported, as well as regions with high environmental suitability for transmission but where the disease has not yet been documented (Fig. 4A) and areas which are environmentally suitable but have low accessibility to healthcare facilities (Fig. 4B). Environmental suitability changes over time The global distribution of environmentally suitable areas for scrub typhus has remained largely stable over the past two decades, with consistently widespread regions of high suitability observed across tropical and subtropical zones (Fig. 5A–C). Despite some localized expansions and contractions of suitable areas between 2001 and 2020 (Fig. 5D–F), the overall extent of environmentally suitable land has remained substantial, with minimal fluctuations in core high-risk regions. Regions in South and Southeast Asia, sub-Saharan Africa, northern Australia, and parts of South America have consistently exhibited high environmental suitability throughout the study period. While minor expansions were observed in areas such as parts of Central Asia and northern South America, these changes did not significantly alter the global distribution pattern. Population exposure has also remained significant, with billions of people residing within environmentally suitable areas over time. Though some shifts in population within suitable areas occurred, the overarching pattern highlights a stable and extensive potential risk zone. Detailed numerical data on changes in suitable area, associated uncertainties, and population exposure across the study period are provided in Supplementary Table S11. Discussion This study presents the first high-resolution global environmental suitability map for scrub typhus, revealing that substantial areas of high suitability extend beyond the traditionally recognized endemic regions. Previous research has largely focused on the Asia-Pacific region 20 , 30 , 52 with only a few descriptive studies addressing the global epidemiology of scrub typhus 53 , 54 . This study build on these efforts by expanding the geographic scope globally. By integrating a comprehensive database of scrub typhus occurrences with an extensive array of environmental covariates, advanced machine learning models were employed to predict areas of high environmental suitability for scrub typhus occurrence. The results underscore the need for a re-evaluation of the global distribution of this neglected tropical disease, which has implications for public health strategies and disease burden estimation. One of the key findings from this analysis is the extensive suitability of environments for scrub typhus across regions that have not historically reported cases. Notably, large areas in Brazil, the United States, Mexico, Nigeria, Ethiopia, and Egypt—countries with no documented local transmission—were identified as potential high-risk zones suitable for scrub typhus occurrence. These findings challenge the traditional confines of the \"tsutsugamushi triangle\" and highlight the potential for scrub typhus to emerge in regions previously considered non-endemic. The expansion of scrub typhus beyond the traditional endemic regions, as evidenced by the recent identification of new Orientia species in Dubai 9 and Chile 10 , the detection of Orientia in field collected free-living Eutrombicula chiggers in the United States 55 and emerging serologic and molecular evidence of Orientia spp. endemicity in Uganda 11 supports the hypothesis that scrub typhus is not confined to the \"tsutsugamushi triangle\". Despite the identification of large environmentally suitable areas for scrub typhus transmission in the Americas and Africa, the number of reported human cases in many countries in these regions remains low or zero. Potential explanations include limited healthcare access, competing national health priorities, historical under-recognition of scrub typhus and the possible absence or low density of suitable reservoir hosts or vector arthropods (e.g., chigger mites) that occasionally bite humans. In the United States, the issue is unlikely to be related to healthcare access or national healthcare expenditure, as most of the population has access to advanced diagnostic facilities. While under-recognition remains a possible factor- given scrub typhus’ nonspecific febrile symptoms, which can easily be mistaken for more common tick-borne diseases such as Lyme disease or spotted fever group rickettsiosis, which are endemic to parts of the U.S 56 , 57 . Additionally, the historical focus of scrub typhus research has been centred on the Asia-Pacific region and it was long assumed that O. tsutsugamushi was restricted to this region. As a result, diagnostic protocols in the U.S. may not routinely include tests for Orientia species unless there is travel history to endemic areas 58 , 59 . Alternative explanations may include the presence of mild or asymptomatic Orientia infections that do not prompt healthcare-seeking behaviour or the existence of local vector species that are unsuitable for efficient human transmission. In contrast, in Africa, healthcare access and national health priorities may play a larger role in the underreporting of scrub typhus. Many regions in Africa face severe challenges in providing even basic healthcare services, with long travel times to healthcare facilities and chronic underfunding of public health systems 60 . Diseases like Ebola, malaria, HIV/AIDS, and tuberculosis, which have higher mortality rates and more severe public health impacts, take priority in resource allocation and healthcare response 61 – 64 . Moreover, the historical relative neglect of scrub typhus scientific research in Africa and the lack of diagnostic tools and awareness among healthcare providers further exacerbates this issue. It is thus possible that scrub typhus is being misdiagnosed and underreported, even though the environmental conditions may support the transmission of Orientia . The potential role of climate change and environmental conditions in driving this expanded range cannot be overlooked. Changes in temperature, precipitation patterns, and habitat availability due to climate change may create new ecological niches for Orientia and its vectors, facilitating the spread of the disease to previously unaffected areas 20 , 22 . In addition, migratory birds may play a critical role in spreading infected mites or vectors across wide geographical areas 65 and contribute to the creation of new hotspots for scrub typhus, particularly in regions where suitable environmental conditions are emerging due to climate change 66 . The relative importance of different covariates in the model, reveals key insights into the factors that most strongly influence the predictive accuracy of scrub typhus environmental suitability. Temperature variables, both maximum temperature and minimum temperature, were shown to have the highest impact on model accuracy, underscoring the critical role of climate in shaping the disease’s distribution. This is consistent with the biology of O. tsutsugamushi and its vectors: warmer and more humid climates tend to favour mite population growth by providing optimal conditions for egg-laying and larvae survival 2 , 67 – 69 . Elevation also ranks highly, as higher elevations tend to be inhospitable for both vector and host survival, with fewer people and lower vegetation cover, creating a natural barrier to scrub typhus transmission 18 , 22 , 29 , 31 , 70 . Socioeconomic factors such as urban accessibility and population density, emerge as significant contributors by elevating the chance of potential human-vector interactions 18 , 31 , 71 . Vegetation density, such as Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI), along with vegetation types, such as evergreen broadleaf forests and croplands, contribute to the ecological niches where both the vector and the pathogen may persist 72 – 74 . Areas with minimal vegetation, like barren lands, seem to have the lowest relative importance, possibly due to lower vector presence or limited human habitation. These associations have been previously recognized in localized studies, but our findings demonstrate their relevance at a global scale, highlighting the importance of incorporating a wide range of covariates when assessing scrub typhus risk. The uncertainties in our environmental suitability estimates, especially in regions with sparse human case occurrence data, underscore the difficulties in accurately and confidently predicting the disease suitability distribution. The widest confidence intervals were observed in areas where there is inconsistent or limited evidence of scrub typhus presence. This was most pronounced in regions where environmental suitability was consistently high across models, yet lacked robust occurrence records to validate predictions. For example, much of Africa, parts of South America, and central Asia exhibited high environmental suitability but suffer from a lack of confirmed cases, leading to increased uncertainty in these areas. Improving data collection efforts in these regions could greatly refine our understanding of the spatial distribution of scrub typhus and, consequently, improve the precision of disease burden estimates. There are several key areas on which these efforts should focus. First, targeted research is needed to confirm the presence of scrub typhus in regions where environmental suitability is high, but occurrence data is lacking. This includes field studies and epidemiological investigations to uncover undetected cases 10 . Second, increasing awareness among physicians and communities in endemic but likely underreported regions is essential for improving diagnosis and reporting 75 , 76 . Enhanced training on recognizing scrub typhus symptoms and the availability of appropriate diagnostic testing, particularly in areas with limited healthcare infrastructure, could significantly reduce underreporting 77 . Engaging communities in health education and sharing research knowledge can further strengthen these efforts, improving overall awareness and encouraging timely healthcare-seeking behaviour 78 . Finally, strengthening national and regional surveillance systems will be crucial for systematically capturing both confirmed and suspected cases. Investments in expanding testing availability and improving reporting mechanisms can lead to a more accurate picture of the disease's global distribution, ultimately enhancing the precision of future models and supporting effective public health interventions. Moreover, this analysis revealed that approximately 2.5 billion people (95% CI: 2.43–2.69 billion) are currently living in environmentally suitable areas within countries or regions where human cases of scrub typhus have already been confirmed. This figure provides the first data-driven estimate of land area and population at risk for scrub typhus, in contrast to the only previously available generalized statement from a 1997 study, which stated that \"about 13 million square kilometres of land are endemic, and more than a billion people would appear to be at risk\" within the tsutsugamushi triangle 79 . This new estimate incorporates not only the traditional areas within the triangle but also countries and regions outside of it where human cases of scrub typhus have been confirmed, or suspected, in recent years. Given the 27 years that have passed since the original estimate, during which global populations have risen and climate has changed substantially, particularly in many endemic countries 80 , this updated figure of 2.5 billion people highlights the growing significance of scrub typhus as a public health threat. Expanding the scope of the analysis globally, it was estimated that over 4.4 billion people live in high environmental suitability areas. While this large number highlights the widespread nature of suitable environments for the disease, it does not imply that large numbers of people are at high risk of infection. Contracting scrub typhus requires several other factors, including exposure to a suitable vector (infected and human-biting), the presence of the disease in the environment, and other ecological and human-mediated pathways 2 . Therefore, while the environmental suitability is high, quantifying the actual risk of infection requires consideration of a complex interplay of additional factors. Despite these extensive efforts to compile comprehensive data on scrub typhus occurrence and employ new modelling approaches to enhance the predictive power of these data, several limitations remain. The empirical evidence base for risk remains limited by the availability and quality of georeferenced occurrence data, which vary significantly across regions. Notably, a substantial portion of the data originate from Asia while significant gaps remain in remote or underreported areas, which may introduce bias when extrapolating findings to other regions with different ecological and epidemiological contexts. The developed model is likely more accurate in well-surveyed/data-rich regions like Southeast Asia, but less reliable in data-sparse areas such as Africa and the Americas. Additionally, while a broad range of global environmental covariates was utilized, they may not capture local variations that could significantly influence disease transmission dynamics. The regional variability in vector and host ecology was not accounted for, which may limit the model's applicability in predicting actual disease risk. Furthermore, underdiagnosis of asymptomatic or mild cases, particularly in areas with limited surveillance, could skew the estimates. These limitations emphasize the need for improved surveillance in underrepresented regions and ongoing research to refine the environmental and biological factors driving scrub typhus transmission, especially in the context of climate change. While this study provides the most comprehensive, evidence-based estimate based on currently available data and methodologies, continued advancements in surveillance, diagnostics, and data collection, alongside the development of new analytical tools, will be essential for refining and improving future assessments. In conclusion, this study provides updated and more in-depth understanding of the global distribution of scrub typhus and highlights the need for more comprehensive surveillance and further investigations. The global burden of scrub typhus could be heavily underestimated and represents a growing challenge to public health officials and policymakers. The high-resolution environmental suitability maps developed in this study offer valuable insights to prioritize regions for further investigation, strengthened surveillance and intervention and to better estimate the global burden of this neglected disease. Success in tackling this growing global threat is, in part, contingent on strengthening the evidence base on which control planning decisions and their impact are evaluated. It is hoped that this evaluation of contemporary environmental suitability will help to advance that goal. Declarations Acknowledgements This research was funded in part by the Wellcome Trust [220211], National Natural Science Foundation of China [42201497] and Youth Innovation Promotion Association [2023000117]. The funder had no role in study design, data collection, data analysis, data interpretation, or writing of this study. For the purpose of Open Access, the author has applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission. We would like to thank Thailand Ministry of Public Health for providing the national surveillance data. All authors read and approved the final manuscript. Contributions RM, BS, and ND conceived the study and provided overall guidance. QW, TM, FD and CZ contributed to data collection. QW had major roles in formulating the analysis under the supervision of RM, BS, and ND. QW prepared the first draft and finalized the manuscript based on feedback from all co-authors. All co-authors have contributed substantially to the review and editing of the manuscript. Ethics declarations Competing interests We declare no competing interests. Data availability The authors had full access to the data and were responsible for conducting the analyses. This research was conducted in compliance with all relevant ethical regulations. Ethical approval is not applicable. These de-identified data may be made available upon reasonable request via a proposal-based process. Interested researchers can contact [email protected] . References Xu, G., Walker, D.H., Jupiter, D., Melby, P.C. & Arcari, C.M. A review of the global epidemiology of scrub typhus. PLoS Neglected Tropical Diseases 11 (2017). Elliott, I. , et al. Scrub typhus ecology: a systematic review of Orientia in vectors and hosts. Parasites & vectors 12 , 513 (2019). Rajapakse, S., Weeratunga, P., Sivayoganathan, S. & Fernando, S.D. Clinical manifestations of scrub typhus. Transactions of the Royal Society of Tropical Medicine and Hygiene 111 , 43-54 (2017). Kelly, D.J., Fuerst, P.A., Ching, W.-M. & Richards, A.L. Scrub typhus: the geographic distribution of phenotypic and genotypic variants of Orientia tsutsugamushi. Clinical infectious diseases 48 , S203-S230 (2009). Fan, M.Y., Walker, D.H., Yu, S.R. & Liu, Q.H. Epidemiology and ecology of rickettsial diseases in the People's Republic of China. Reviews of infectious diseases 9 , 823-840 (1987). Philip, C.B. Observations on Tsutsugamushi Disease (Mite-borne or Scrub Typhus) in northwest Honshu Island, Japan, in the Fall of 1945. I. Epidemiological and ecological Data. American Journal of Hygiene 46 , 45-pp (1947). Wangrangsimakul, T. , et al. The estimated burden of scrub typhus in Thailand from national surveillance data (2003-2018). PLoS Neglected Tropical Diseases 14 , 1-20 (2020). Devasagayam, E. , et al. The burden of scrub typhus in India: A systematic review. PLoS neglected tropical diseases 15 , e0009619 (2021). Izzard, L. , et al. Isolation of a novel Orientia species (O. chuto sp. nov.) from a patient infected in Dubai. J Clin Microbiol 48 , 4404-4409 (2010). Weitzel, T. , et al. Endemic Scrub Typhus in South America. The New England journal of medicine 375 , 954-961 (2016). Blair, P.W. , et al. Evidence of Orientia spp. Endemicity among Severe Infectious Disease Cohorts, Uganda. Emerg Infect Dis 30 , 1442-1446 (2024). Izzard, L. , et al. Isolation of a novel Orientia species (O. chuto sp. nov.) from a patient infected in Dubai. Journal of clinical microbiology 48 , 4404-4409 (2010). Abarca, K. , et al. Molecular Description of a Novel Orientia Species Causing Scrub Typhus in Chile. Emerg Infect Dis 26 , 2148-2156 (2020). Wangrangsimakul, T. , et al. The estimated burden of scrub typhus in Thailand from national surveillance data(2003-2018). PLoS neglected tropical diseases 14 , e0008233 (2020). Li, Z. , et al. Epidemiologic Changes of Scrub Typhus in China, 1952-2016. Emerging infectious diseases 26 , 1091-1101 (2020). Yu, H. , et al. Scrub typhus in Jiangsu Province, China: epidemiologic features and spatial risk analysis. BMC Infectious Diseases 18 , 372 (2018). Acharya, B.K. , et al. Mapping Environmental Suitability of Scrub Typhus in Nepal Using MaxEnt and Random Forest Models. International journal of environmental research and public health 16 (2019). Xin, H. , et al. Risk mapping of scrub typhus infections in Qingdao city, China. PLoS neglected tropical diseases 14 , e0008757 (2020). Wang, Q. , et al. A systematic review of environmental covariates and methods for spatial or temporal scrub typhus distribution prediction. Environ Res 263 , 120067 (2024). Ding, F. , et al. Climate drives the spatiotemporal dynamics of scrub typhus in China. Glob Chang Biol 28 , 6618-6628 (2022). Ogawa, T. , et al. Analysis of Differences in Characteristics of High-Risk Endemic Areas for Contracting Japanese Spotted Fever, Tsutsugamushi Disease, and Severe Fever With Thrombocytopenia Syndrome. Open Forum Infect Dis 11 , ofae025 (2024). Roberts, T. , et al. A spatio-temporal analysis of scrub typhus and murine typhus in Laos; implications from changing landscapes and climate. PLoS neglected tropical diseases 15 , e0009685 (2021). J, Q. , et al. Spatiotemporal heterogeneity and long-term impact of meteorological, environmental, and socio-economic factors on scrub typhus in China from 2006 to 2018. BMC public health 24 , 538 (2024). Chang, T., Min, K.D., Cho, S.I. & Kim, Y. Associations of meteorological factors and dynamics of scrub typhus incidence in South Korea: A nationwide time-series study. Environ Res 245 , 117994 (2024). Mungmungpuntipantip, R. & Wiwanitkit, V. Correlation between rainfall and the prevalence of scrub typhus: an observation from a tropical endemic country. Int.j.med.surg.sci.(Print) 8 , 1-4 (2021). 吴义城, 张文义 & 李申龙. 中国人民解放军军事医学科学院 (2016). Wei, Y. , et al. Climate variability, animal reservoir and transmission of scrub typhus in Southern China. PLoS Neglected Tropical Diseases 11 , e0005447 (2017). Lu, J., Liu, Y., Ma, X., Li, M. & Yang, Z. Impact of Meteorological Factors and Southern Oscillation Index on Scrub Typhus Incidence in Guangzhou, Southern China, 2006-2018. Front Med (Lausanne) 8 , 667549 (2021). Min, K.-D., Lee, J.-Y., So, Y. & Cho, S.-I. Deforestation Increases the Risk of Scrub Typhus in Korea. International journal of environmental research and public health 16 (2019). Wangrangsimakul, T. , et al. The estimated burden of scrub typhus in Thailand from national surveillance data (2003-2018). PLoS Negl Trop Dis 14 , e0008233 (2020). Zheng, C., Jiang, D., Ding, F., Fu, J. & Hao, M. Spatiotemporal Patterns and Risk Factors for Scrub Typhus From 2007 to 2017 in Southern China. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 69 , 1205-1211 (2019). 刘晓宁. 硕士, 安徽医科大学 (2019). Huang, X. , et al. Prediction of risk factors for scrub typhus from 2006 to 2019 based on random forest model in Guangzhou, China. Tropical Medicine & International Health 28 , 551-561 (2023). Stevens, G.A. , et al. Guidelines for Accurate and Transparent Health Estimates Reporting: the GATHER statement. The Lancet 388 , e19-e23 (2016). Wang, Q. , et al. Global and regional seroprevalence, incidence, mortality of, and risk factors for scrub typhus: A systematic review and meta-analysis. Int J Infect Dis 146 , 107151 (2024). Araújo, M.B. & New, M. Ensemble forecasting of species distributions. Trends Ecol Evol 22 , 42-47 (2007). Hastie, T.J. Generalized additive models. in Statistical models in S 249-307 (Routledge, 2017). Elith, J., Leathwick, J.R. & Hastie, T. A working guide to boosted regression trees. Journal of animal ecology 77 , 802-813 (2008). Breiman, L. Random forests. Machine learning 45 , 5-32 (2001). Prasad, A.M., Iverson, L.R. & Liaw, A. Newer classification and regression tree techniques: bagging and random forests for ecological prediction. Ecosystems 9 , 181-199 (2006). Tibshirani, R. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society Series B: Statistical Methodology 58 , 267-288 (1996). Nelder, J.A. & Wedderburn, R.W. Generalized linear models. Journal of the Royal Statistical Society Series A: Statistics in Society 135 , 370-384 (1972). Hastie, T., Tibshirani, R., Friedman, J.H. & Friedman, J.H. The elements of statistical learning: data mining, inference, and prediction , (Springer, 2009). Bhatt, S. , et al. Improved prediction accuracy for disease risk mapping using Gaussian process stacked generalization. J R Soc Interface 14 (2017). Zhou, Z.-H. Ensemble methods: foundations and algorithms , (CRC press, 2025). Wolpert, D.H. Stacked generalization. Neural networks 5 , 241-259 (1992). Wilson, E.B. Probable Inference, the Law of Succession, and Statistical Inference. Journal of the American Statistical Association 22 , 209-212 (1927). Efron, B. & Tibshirani, R.J. An introduction to the bootstrap , (Chapman and Hall/CRC, 1994). Phillips, S.J. , et al. Sample selection bias and presence‐only distribution models: implications for background and pseudo‐absence data. Ecological applications 19 , 181-197 (2009). Williams, J.N. , et al. Using species distribution models to predict new occurrences for rare plants. Diversity and distributions 15 , 565-576 (2009). Weiss, D.J. , et al. Global maps of travel time to healthcare facilities. Nature Medicine 26 , 1835-1838 (2020). Devasagayam, E. , et al. The burden of scrub typhus in India: A systematic review. PLoS Negl Trop Dis 15 , e0009619 (2021). Bonell, A., Lubell, Y., Newton, P.N., Crump, J.A. & Paris, D.H. Estimating the burden of scrub typhus: A systematic review. PLoS Negl Trop Dis 11 , e0005838 (2017). Xu, G., Walker, D.H., Jupiter, D., Melby, P.C. & Arcari, C.M. A review of the global epidemiology of scrub typhus. PLoS Negl Trop Dis 11 , e0006062 (2017). Chen, K. , et al. Detection of Orientia spp. Bacteria in Field-Collected Free-Living Eutrombicula Chigger Mites, United States. Emerg Infect Dis 29 , 1676-1679 (2023). Schwartz, A.M. Surveillance for lyme disease—United States, 2008–2015. MMWR. Surveillance Summaries 66 (2017). Biggs, H.M. Diagnosis and management of tickborne rickettsial diseases: Rocky Mountain spotted fever and other spotted fever group rickettsioses, ehrlichioses, and anaplasmosis—United States. MMWR. Recommendations and Reports 65 (2016). Hendershot, E.F. & Sexton, D.J. Scrub typhus and rickettsial diseases in international travelers: a review. Current infectious disease reports 11 , 66-72 (2009). Sultana, R. , et al. The Brief Case: A traveler’s tale—imported scrub typhus in a child returning from Bangladesh. Journal of Clinical Microbiology 61 , e00359-00323 (2023). Weiss, D. , et al. Global maps of travel time to healthcare facilities. Nature medicine 26 , 1835-1838 (2020). Team, W.E.R. Ebola virus disease in West Africa—the first 9 months of the epidemic and forward projections. New England Journal of Medicine 371 , 1481-1495 (2014). Bhatt, S. , et al. The effect of malaria control on Plasmodium falciparum in Africa between 2000 and 2015. Nature 526 , 207-211 (2015). Hotez, P.J. & Kamath, A. Neglected tropical diseases in sub-Saharan Africa: review of their prevalence, distribution, and disease burden. PLoS neglected tropical diseases 3 , e412 (2009). Gandhi, N.R. , et al. Extensively drug-resistant tuberculosis as a cause of death in patients co-infected with tuberculosis and HIV in a rural area of South Africa. The Lancet 368 , 1575-1580 (2006). Scott, J.D. Studies abound on how far north Ixodes scapularis ticks are transported by birds. Ticks and tick-borne diseases 7 , 327-328 (2016). Walker, D.H. Scrub Typhus — Scientific Neglect, Ever-Widening Impact. New England Journal of Medicine 375 , 913-915 (2016). Audy, J.R. The ecology of scrub typhus. Studies in disease ecology 2 , 389-432 (1961). Wharton, G.W. & Fuller, H.S. A Manual of the Chiggers. The Biology, Classification, Distribution, and Importance to Man of the Larvae of the Family Trombiculidae (Acariña). (1952). Kawamura, R. & Ikeda, K. Ecological Study of the Tsutsugamushi, Trombicula akamushi (Brumpt). (1936). L, L. , et al. Spatiotemporal epidemiology and risk factors of scrub typhus in Hainan Province, China, 2011-2020. One health (Amsterdam, Netherlands) 17 , 100645 (2023). 孙烨, 方., 曹务春. 中国人民解放军军事医学科学院 (2016). Traub, R. & Wisseman Jr, C.L. The ecology of chigger-borne rickettsiosis (scrub typhus). Journal of medical entomology 11 , 237-303 (1974). Santibáñez, P., Palomar, A.M., Portillo, A., Santibáñez, S. & Oteo, J.A. The role of chiggers as human pathogens. An overview of tropical diseases 1 , 173-202 (2015). Chaisiri, K., Cosson, J.-F. & Morand, S. Infection of rodents by Orientia tsutsugamushi, the agent of scrub typhus, in relation to land use in Thailand. Tropical Medicine and Infectious Disease 2 , 53 (2017). Xu, G., Walker, D.H., Jupiter, D., Melby, P.C. & Arcari, C.M. A review of the global epidemiology of scrub typhus. PLoS neglected tropical diseases 11 , e0006062 (2017). Blacksell, S.D. , et al. Underrecognized arthropod-borne and zoonotic pathogens in northern and northwestern Thailand: serological evidence and opportunities for awareness. Vector borne and zoonotic diseases (Larchmont, N.Y.) 15 , 285-290 (2015). Sharma, R. Scrub typhus: prevention and control. JK science 12 , 91 (2010). Perrone, C. , et al. Community engagement around scrub typhus in northern Thailand: a pilot project. Transactions of The Royal Society of Tropical Medicine and Hygiene 118 , 666-673 (2024). Rosenberg, R. Drug-resistant scrub typhus: Paradigm and paradox. Parasitol Today 13 , 131-132 (1997). Mohapatra, R.K. , et al. Linking the increasing epidemiology of scrub typhus transmission in India and South Asia: Are the varying environment and the reservoir animals the factors behind? Frontiers in Tropical Diseases 5 , 1371905 (2024). Additional Declarations There is NO Competing Interest. Supplementary Files SI18Mar25.pdf Supplementary Material Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-6251403\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Article\",\"associatedPublications\":[],\"authors\":[{\"id\":431538757,\"identity\":\"ff30d51d-d6a5-4028-afbb-aa0c63290ff0\",\"order_by\":0,\"name\":\"Richard Maude\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6ElEQVRIiWNgGAWjYDACZgY2MAkEBgwfgCQbOylaGGeAtDATtgehhZmHAcbGA8zbmZ89+LnHOpp/RvLGxza/tsnzMTMwfviYg1uLzGE2c8OeZ+m5M26kFRvn9t02bGNmYJacuQ23FglmHjYJngOHcxtu5JhJ5/bcZgRqYWPmJaBF8g9Qy/wbOea/LXtu2xOlRRpkywagLcwMP24nEqGFzUxa5kB67sYzz4olextuJ7cxMzbj9wv/4WeSbw5Y5847nrzxw48/t23ntzcf/PARjxZUwNgGJhuIVQ8Cf0hRPApGwSgYBSMFAABOnkzEyYCHiAAAAABJRU5ErkJggg==\",\"orcid\":\"https://orcid.org/0000-0002-5355-0562\",\"institution\":\"Mahidol Oxford Tropical Medicine Research Unit\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Richard\",\"middleName\":\"\",\"lastName\":\"Maude\",\"suffix\":\"\"},{\"id\":431538758,\"identity\":\"d5f0ca70-c672-4449-8fef-321ab97e6940\",\"order_by\":1,\"name\":\"Qian Wang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Mahidol Oxford Tropical Medicine Research Unit, Faculty of Tropical Medicine, Mahidol University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Qian\",\"middleName\":\"\",\"lastName\":\"Wang\",\"suffix\":\"\"},{\"id\":431538759,\"identity\":\"fc5bfde5-c94d-4743-acdb-328eeb253c0b\",\"order_by\":2,\"name\":\"Tian Ma\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Yale Institute for Biospheric Studies, Yale University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Tian\",\"middleName\":\"\",\"lastName\":\"Ma\",\"suffix\":\"\"},{\"id\":431538760,\"identity\":\"5f535a52-6887-41f6-b0c6-445301b6a009\",\"order_by\":3,\"name\":\"Fangyu Ding\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Fangyu\",\"middleName\":\"\",\"lastName\":\"Ding\",\"suffix\":\"\"},{\"id\":431538761,\"identity\":\"3323b701-dad9-4801-aa76-4e6b49a3bc83\",\"order_by\":4,\"name\":\"Ivo Elliott\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ivo\",\"middleName\":\"\",\"lastName\":\"Elliott\",\"suffix\":\"\"},{\"id\":431538762,\"identity\":\"08f1a1cf-f1dd-492b-99eb-0fe5680b333e\",\"order_by\":5,\"name\":\"Canjun Zheng\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Chinese Center for Disease Control and Prevention\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Canjun\",\"middleName\":\"\",\"lastName\":\"Zheng\",\"suffix\":\"\"},{\"id\":431538763,\"identity\":\"ef663cac-7503-402f-b9e4-f0811322ad21\",\"order_by\":6,\"name\":\"Nicholas P. Day\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of Oxford\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Nicholas\",\"middleName\":\"P.\",\"lastName\":\"Day\",\"suffix\":\"\"},{\"id\":431538764,\"identity\":\"45f47fd6-3b66-4b02-b266-eca9feb35119\",\"order_by\":7,\"name\":\"Benn Sartorius\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0001-6761-2325\",\"institution\":\"University of Queensland\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Benn\",\"middleName\":\"\",\"lastName\":\"Sartorius\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-03-18 09:12:28\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-6251403/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-6251403/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":79216871,\"identity\":\"66ab7904-97d0-4a80-93ac-44b1a6cba733\",\"added_by\":\"auto\",\"created_at\":\"2025-03-25 19:01:04\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":281042,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003ePredicted environmental suitability for scrub typhus at 5 × 5 km\\u003csup\\u003e2\\u003c/sup\\u003e spatial resolution (Central DR Congo data gap due to missing population density data).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6251403/v1/30e3213ff844c20cdfd62b19.png\"},{\"id\":79216870,\"identity\":\"f96f67b8-2d13-488c-b0ff-95b168a9b127\",\"added_by\":\"auto\",\"created_at\":\"2025-03-25 19:01:04\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":61376,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eComparison of reported incidence and population environmental suitability. (A) Comparison of mean annual incidence in administrative level 1 areas with low and high environmental suitability across Mainland China (provinces; n=31), South Korea (provincial-level divisions; n=17) and Japan (prefectural divisions; n=47); (B) GLM regression analysis of reported incidence and population weighted environmental suitability across Mainland China, South Korea and Japan.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6251403/v1/4c1e0811c906c98dd8389ed7.png\"},{\"id\":79216873,\"identity\":\"c73b33ba-73b4-4766-bd5f-652dfee64e9e\",\"added_by\":\"auto\",\"created_at\":\"2025-03-25 19:01:05\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":283569,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eOverlapping environmental suitability and width of the confidence interval at 5x5-km grid cell level (Central DR Congo data gap due to missing population density data).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6251403/v1/0e154d2e623153840cbf7132.png\"},{\"id\":79217492,\"identity\":\"3a4251e6-3cd4-4769-9239-32470caefe19\",\"added_by\":\"auto\",\"created_at\":\"2025-03-25 19:09:05\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":236605,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003ePriority countries/regions where targeted research and enhanced surveillance are most urgently needed. (A) Countries/regions where population weighted environmental suitability \\u0026gt; 0.5 but no confirmed cases in humans have been reported; (B) Areas where population weighted environmental suitability \\u0026gt; 0.5 which have low accessibility to healthcare facilities (travel time to healthcare facilities over 1hr \\u003csup\\u003e51\\u003c/sup\\u003e).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6251403/v1/85bd75b085a925f7596da289.png\"},{\"id\":79216876,\"identity\":\"a5556caa-8fac-4bf3-9e50-6a4637a7900f\",\"added_by\":\"auto\",\"created_at\":\"2025-03-25 19:01:05\",\"extension\":\"png\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":483613,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eTemporal change of predicted global environmental suitability for scrub typhus. (A-C) Predicted global environmental suitability of 2001, 2010 and 2020, (D) Expansion and contraction of environmentally suitable area from 2001 to 2010, (E) Expansion and contraction of environmentally suitable area from 2010 to 2020, (F) Expansion and contraction of environmentally suitable area from 2001 to 2020.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6251403/v1/b833e0f477883e820ad84efb.png\"},{\"id\":80332184,\"identity\":\"907dc873-fc85-4d26-8b0e-d2c0bca07382\",\"added_by\":\"auto\",\"created_at\":\"2025-04-10 15:33:44\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":1983313,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6251403/v1/c2b44f21-ab66-41cc-a68c-020405768534.pdf\"},{\"id\":79216874,\"identity\":\"be6df5de-aa98-425f-9eab-674d1e07f7bf\",\"added_by\":\"auto\",\"created_at\":\"2025-03-25 19:01:05\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1587555,\"visible\":true,\"origin\":\"\",\"legend\":\"Supplementary Material\",\"description\":\"\",\"filename\":\"SI18Mar25.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6251403/v1/4bffb508c6f2f9dac4249b33.pdf\"}],\"financialInterests\":\"There is \\u003cb\\u003eNO\\u003c/b\\u003e Competing Interest.\",\"formattedTitle\":\"Mapping the global distribution of and environmental suitability for scrub typhus\",\"fulltext\":[{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eScrub typhus, also known as tsutsugamushi disease, is an acute febrile illness and a neglected vector-borne infectious disease\\u0026nbsp;that poses a serious public health threat globally\\u003csup\\u003e1\\u003c/sup\\u003e. It is mainly caused by the bacterium \\u003cem\\u003eOrientia tsutsugamushi\\u003c/em\\u003e and is transmitted to humans through the bite of the larval stage of trombiculid mites, “chiggers”\\u003csup\\u003e2\\u003c/sup\\u003e. The disease presents with symptoms ranging from fever, headache, and rash to more severe complications such as pneumonitis, meningoencephalitis, and multi-organ failure\\u003csup\\u003e3\\u003c/sup\\u003e, making it a critical public health concern.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eOriginally, the endemic area for scrub typhus was considered to be restricted to the geographic “tsutsugamushi triangle” which extends from the far east of Russia in the north, to northern Australia in the south, and Afghanistan and Pakistan in the west.\\u003csup\\u003e4\\u003c/sup\\u003e Many countries in this region have suffered from scrub typhus for a long time, particularly China, Japan, Korea, Thailand, and India\\u003csup\\u003e5-8\\u003c/sup\\u003e.\\u0026nbsp;But there is mounting evidence to suggest that the range of scrub typhus expands beyond these traditional boundaries to the Middle East, South America, and Africa\\u003csup\\u003e9-11\\u003c/sup\\u003e. Recently, two new \\u003cem\\u003eOrientia\\u003c/em\\u003e species causing scrub typhus were isolated in unexpected areas: \\u003cem\\u003eCandidatus\\u003c/em\\u003e Orientia chuto in Dubai\\u003csup\\u003e12\\u003c/sup\\u003e and \\u003cem\\u003eCandidatus\\u0026nbsp;\\u003c/em\\u003eOrientia chiloensisin Chile\\u003csup\\u003e13\\u003c/sup\\u003e, which not only indicates a geographic expansion, but may also imply a lack of awareness and attention to the disease outside of the triangle.\\u0026nbsp;Detailed spatial information on the global geographical extent of scrub typhus and areas that are environmentally suitable for transmission to humans remains sparse. Previous studies have primarily assessed scrub typhus at national level\\u003csup\\u003e14,15\\u003c/sup\\u003e, with risk maps generated only for a few countries or subnational units\\u003csup\\u003e16\\u003c/sup\\u003e\\u003csup\\u003e,\\u003c/sup\\u003e\\u003csup\\u003e17,18\\u003c/sup\\u003e. Given recent improvement in empirical data as well as advances in disease modelling approaches, a comprehensive assessment of scrub typhus records and the creation of high-resolution environmental suitability maps on a global scale was carried out.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eCurrently, spatial analyses of scrub typhus are few in number compared to some other neglected tropical diseases (NTDs). Several studies have attempted to identify the chief determinants of scrub typhus geographic distribution patterns\\u003csup\\u003e19-25\\u003c/sup\\u003e. Strong correlations have been found in China, Thailand, South Korea, Japan and Nepal between scrub typhus occurrence and suitable environments, including meteorological factors, geographic factors, socioeconomic variables, and biological factors. Temperature, precipitation, and humidity, as well as the El Niño/Southern Oscillation (ENSO) and Multivariate ENSO Index, have been found to be associated with scrub typhus in China\\u003csup\\u003e26-28\\u003c/sup\\u003e. Geographic indicators have included elevation, Normalized Difference Vegetation Index (NDVI), cropland, forest, and grassland landcovers. Deforestation exhibited positive associations with scrub typhus incidence in South Korea. One study in Thailand showed habitat complexity and fragmentation to be significant predictors for scrub typhus cases\\u003csup\\u003e29,30\\u003c/sup\\u003e.\\u0026nbsp;The numbers of patients with scrub typhus infection were found to be higher in villages with higher levels of surface flooding and vegetation in a study in Laos\\u003csup\\u003e22\\u003c/sup\\u003e. Socioeconomic features have also been associated with spatial patterns of scrub typhus risk, such as a negative association between travel time to cities and reported cases in China\\u003csup\\u003e31\\u003c/sup\\u003e.\\u0026nbsp;Biological factors, such as rodent density and Orientia infection rate in mites/rodent also have been examined to have association with scrub typhus\\u003csup\\u003e27,32,33\\u003c/sup\\u003e.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eKnowledge of the geographical distribution of potential scrub typhus infection risk is critical for identifying and prioritizing regions for preventive and targeted control measures and is essential for estimating the global clinical burden of scrub typhus, which is currently unknown. Therefore, a comprehensive database of timepoints and locations where scrub typhus has been reported was developed. Then these data were coupled with a comprehensive set of environmental covariates to generate high-resolution risk maps (5km x 5km), representing global environmental suitability for scrub typhus occurrence. Three single machine learning models and two stacked ensemble methods were applied to generate these risk maps with 95% uncertainty intervals. By developing a global environmental suitability map, the aim was to provide insights into the potential spatial distribution of scrub typhus risk and identify potential hotspots for further investigation, targeted surveillance and control measures. The results of this study will contribute to the development of evidence-based strategies to mitigate the impact of this neglected tropical disease.\\u003c/p\\u003e\"},{\"header\":\"Methods\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eOverview/Summary\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eA comprehensive global database of geo-located scrub typhus occurrence records was compiled, integrating data from published literature and national surveillance systems, and standardized to annual intervals. Absence data were rigorously verified. Environmental and socio-economic covariates were processed to a unified 5 x 5 km resolution. An ensemble machine learning approach, including generalized additive models, boosted regression trees, and random forests, was conducted combining multiple models to predict the probability of occurrence (with uncertainty) as a measure of environmental suitability. The population living in suitable area was estimated by spatially linking the predicted environmental suitability with high-resolution human population data, enabling an assessment of populations potentially exposed to scrub typhus. The GATHER(Guidelines for Accurate and Transparent Health Estimates Reporting) checklist was followed to report the description of input data and estimate method\\u003csup\\u003e34\\u003c/sup\\u003e (Supplementary Table S1).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eThe occurrence database\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eA database of 67,905 geo-located occurrence records was initially compiled for the period 1944 to 2024, drawing on a diverse array of sources including published literature, online repositories, and national surveillance systems.\\u0026nbsp;After thorough standardization, quality control, and deduplication, the final dataset comprised 56,093 unique year-location occurrences, serving as the basis for modelling.\\u003c/p\\u003e\\n\\u003cp\\u003eA systematic literature search was conducted to capture all relevant occurrences, irrespective of language, publication date, or geographical focus. The processes of searching, screening, data extraction, cleaning, and geolocation have been thoroughly described in our earlier published work\\u003csup\\u003e35\\u003c/sup\\u003e. The occurrence database was first created from published literature, case reports/series and grey literatures and was last updated in May 2024. This extensive review yielded data from 829 published references, resulting in the extraction of 7,165 unique year-location occurrence datapoints, which were subsequently subjected to temporal standardization and spatial processing.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eAs scrub typhus remains a notifiable disease in several Asia countries/regions, data was also sourced from national surveillance systems. We retrieved official records from the online public databases of Japan, South Korea, and Taiwan, and secured access to unpublished national data from Mainland China and Thailand through direct correspondence with relevant health authorities. Detailed methodologies pertaining to data acquisition and processing from these sources are provided in the Supplementary Information section 1.1.\\u003c/p\\u003e\\n\\u003cp\\u003eAll records were standardized annually by location, meaning that repeated records from the same location within a single year were consolidated into a singular occurrence and underwent rigorous quality control, as outlined in Supplementary Information section 1.2. The final occurrence database contained 56,093 unique year-location occurrences, which represent a unique location where one or more cases occurred within one year. A map of the final set of occurrence locations utilized for modelling the contemporary distribution of risk for scrub typhus is provided in Supplementary Fig. S1(a), with the number of occurrences per year globally and by country/region is shown in Supplementary Fig. S1(b).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eBackground location database\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eOur methods require both presence and absence data to define areas of disease absence and potentially unsuitable environmental conditions at unsampled locations. To ensure the absence data is as accurate as possible, it was sourced from both published literature and national surveillance systems. The negative records were extracted from the published sources and cross checked their geolocations with other published literature and national surveillance data by applying dual approach tailored to the nature of the original data. For point-based data, a 30 km radius was applied to ensure no positive or reported cases were present within the surrounding area. For area-based data, the verification was conducted within the same administrative polygons to confirm the absence of reported cases. Additionally, a significant proportion of absence data came from national official reports. Multiple years of data were examined for the smallest administrative divisions, retaining units that had never reported any cases, whether suspected probable, or confirmed. These locations were also cross-checked with literature records to ensure no positive cases were documented. The final absence data comprised 122 published literature records from 13 countries/regions, 400,829 reported absence year-county records from the China CDC, 476 reported year-district records from the Taiwan CDC, and 384 reported year-district records from the Thailand Division of Vector Borne Diseases (DVBD). To balance the presence-absence ratio for modelling, a subset of absence records was randomly selected from this database, ensuring a proportionate distribution relative to the 56,093 initial presence records.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eCovariates\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eGiven that humans are typically incidental hosts and that the bacterium causing scrub typhus is primarily maintained in, and transmitted by, chigger mites within natural cycles, the activities and distribution of these vectors, the presence, movements and population fluctuations of small mammal hosts—particularly rodents and human behaviours, are inherently linked to the occurrence of the disease. These factors are strongly influenced by climatic and other environmental conditions. To identify indicators that have been shown previously to have significant associations with scrub typhus, a systematic review was conducted to extract existing evidence\\u003csup\\u003e19\\u003c/sup\\u003e. This systematic review, which analysed data from 68 articles published between 1978 and 2024 across seven countries/regions, identified 68 significant environmental risk factors associated with scrub typhus with temperature, precipitation, humidity, sunshine duration, elevation, the normalized difference vegetation index (NDVI), the proportion of cropland, population density, and urban status as the top-ten indicators most mentioned.\\u0026nbsp;Based on those findings and the availability of high-resolution spatial and temporal data, 28 covariates were selected for further analysis encompassing meteorological, geographic, and socioeconomic factors.\\u0026nbsp;Meteorological factors included minimum temperature, maximum temperature, accumulated precipitation, relative humidity, surface air pressure, and wind speed. Geographical predictors included the proportions of 17 land cover classes, NDVI, Enhanced Vegetation Index (EVI), and elevation. Socio-economic covariates included travel time to major cities (urban accessibility) and population density. Detailed descriptions and sources of these covariates can be found in the Supplementary Information section 2 and Table S4. All raster data were aggregated and resampled to a unified spatial resolution of 5 km x 5 km and a yearly temporal resolution (Figure S2).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cem\\u003eSpatial processin\\u003c/em\\u003eg Various data formats were converted and integrated into raster format Where necessary, input data sources were re-projected using a standardized equirectangular Plate Carrée projection under the World Geodetic System 1984 (WGS 84) coordinate system. Input grids with spatial resolutions different from 5 km x 5 km were either aggregated or disaggregated to this target resolution using bilinear or nearest-neighbour interpolation techniques.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cem\\u003eTemporal processing\\u0026nbsp;\\u003c/em\\u003eAll data were standardized to annual intervals. Temporal interpolation was not required for most covariates. Elevation and urban accessibility were treated as static covariates, with elevation considered constant over time and urban accessibility data was available only for the year 2015.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cem\\u003eData merging\\u003c/em\\u003e For polygon-based occurrence data from surveillance systems, the mean covariate values within each polygon were calculated. For literature-reported occurrences recorded as hospital locations, the mean covariate value within a 30 km buffer zone around the hospital was calculated. This buffer distance was determined by analysing the distance between reported hospitals and onset locations from national surveillance systems across five countries/regions. For point location data, the covariate values at the exact location were used. All these processes were carried out using ArcGIS (10.8.2) and Python (3.9.13), utilizing the following packages: pandas, geopandas, rasterio, shapely, rasterstats, pyproj, and pygrib.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eModel\\u0026nbsp;\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAn ensemble machine learning approach\\u003csup\\u003e36\\u003c/sup\\u003e was implemented to predict the environmental suitability for scrub typhus based on environmental conditions sampled at each site from the covariate suite, capturing potential nonlinear effects and complex interactions among covariates. After conducting multiple rounds of experimentation, Generalized Additive Models (GAM)\\u003csup\\u003e37\\u003c/sup\\u003e, Boosted Regression Trees (BRT)\\u003csup\\u003e38\\u003c/sup\\u003e, and Random Forest (RF)\\u003csup\\u003e39,40\\u003c/sup\\u003e were selected due to their superior performance as the three primary child models to fit the occurrence data with the covariates as predictors, as measured by the Area Under the Curve (AUC), compared to other models such as Lasso regression\\u003csup\\u003e41\\u003c/sup\\u003e and Generalized Linear Models (GLM)\\u003csup\\u003e42\\u003c/sup\\u003e. Rigorous and comprehensive model selection and parameter tuning were performed on each model to compare among a number of alternative models and enhance the performance. For GAM, covariates exhibiting high concurvity, a condition that generalizes co-linearity and can complicate model interpretation, were subsequently removed. To further enhance model robustness and mitigate the risk of overfitting, a backward stepwise procedure was employed for covariate selection.\\u0026nbsp;For BRT and RF, a balanced dataset approach was adopted to ensure equal representation of both presence and absence samples. This strategy was particularly important for preventing bias in the model’s predictions when dealing with imbalanced datasets. Additionally, extensive parameter tuning was conducted for each of these machine-learning methods, involving a thorough search for optimal hyperparameters (n.trees, shrinkage, interaction.depth, bag.fraction, n.minobsinnode,\\u0026nbsp;mtry) to maximize model performance and ensure generalization across different scenarios. All these models were carried out using R, utilizing the following packages: sf, mgcv, gbm, caret, randomForest. parallel, doParallel. The meticulous selection processes are detailed in the Supplementary Information, section 3.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e70% of the whole dataset was used to train these sub-models and 30% of the data was used for out of sample validation. Each sub-model was fitted using fivefold cross-validation to avoid overfitting, and the out-of-sample predictions from the five folds were compiled into a single set of predictions that were used to fit two stacking methods: constrained weighted mean (CWM) and weighted mean based on R-square weighted mean (RWM)\\u003csup\\u003e43-46\\u003c/sup\\u003e.\\u0026nbsp;The CWM is a stacking ensemble technique that assigns weights to base models within predefined constraints to ensure stability and interpretability, balancing model complexity and predictive accuracy. The RWM assigns weights proportional to each model's R-squared value, emphasizing explanatory power. Both methods optimize the contribution of individual models, enhancing the overall performance and robustness of the ensemble. In addition, each sub-model was also fitted to the full dataset to generate a complete set of in-sample predictions that were subsequently used when generating predictions from stacked ensemble models. The mean area under the receiver operating characteristic curve (AUC) statistics among training data and validation data were used to compare the goodness of fit of those three sub-models and two stacking methods. The predicted environmental suitability was on a scale from 0 to 1, with a final prediction map generated using the best-performing model.\\u003c/p\\u003e\\n\\u003cp\\u003eFor the uncertainty analysis, the Wilson Score interval method was employed to quantify uncertainty and calculate the 95% confidence interval\\u003csup\\u003e47\\u003c/sup\\u003e. To rigorously assess the robustness of the models, three strategies were implemented: (1) Bootstrap Consistency Check: a bootstrap approach\\u003csup\\u003e48\\u003c/sup\\u003e, where 100 distinct training datasets were generated, machine learning models executed iteratively across these datasets, and the variance meticulously analysed among the 100 predictive outputs and with the final prediction; (2)\\u0026nbsp;No-Time-Redundancy Prediction: integration of temporal information by ensuring that each geographic location, irrespective of its temporal reporting frequency, was uniquely represented in the model, thereby removing temporal redundancy. The adjusted dataset was then fitted using the models to generate the no-time-redundancy prediction result; and (3) Absence-Adjusted Prediction:\\u0026nbsp;modification of the presence-to-absence ratio, testing the model's resilience under different conditions with ratios of 1:1, 1:2, and 1:5\\u003csup\\u003e49\\u003c/sup\\u003e. These ratios reflect different scenarios of class imbalance, with 1:1 representing equal numbers of presence and absence data points, while 1:2 and 1:5 progressively increase the proportion of absence data to presence data. The rationale behind testing these ratios was to evaluate the model's stability and predictive power under varying degrees of absence data, as absence data often outnumbers presence data in ecological and disease models\\u003csup\\u003e50\\u003c/sup\\u003e. The model was refitted for each adjusted ratio to generate the absence-adjusted prediction result and assess predictive stability. The stability and robustness of the predictive outcomes were further evaluated using standard deviation metrics and corresponding maps, facilitating a comprehensive comparison of prediction variations across different scenarios.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003eAreas were classified as \\\"at-risk\\\" based on a threshold suitability of 0.5 or greater applied to continuous suitability map values. \\u0026nbsp;Specifically, any pixel with a predicted suitability probability value greater than 0.5 was designated as belonging to an at-risk area. This same threshold was subsequently used to calculate the population within the delineated at-risk areas. Additionally, the areas and populations at risk were further refined by calculating confidence intervals based on the 95% confidence intervals (CI) of the environmental suitability results, providing a more robust estimate of the uncertainty associated with these predictions.\\u003c/p\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cp\\u003eA total of 78,532 georeferenced occurrence and absence year-locations were used to fit the models and 33,656 year-locations to validate them, of which 56,093 were occurrence records and an equal number of background points. Based on the AUC from various test methods (Table S6), the RF method was chosen for the final prediction, and subsequent testing demonstrated good robustness of the final model (Table S8 and Figure S7).\\u003c/p\\u003e\\n\\u003ch3\\u003eRelative importance of covariates\\u003c/h3\\u003e\\n\\u003cp\\u003eBased on the normalized relative importance of various covariates as determined by the model using the Mean Decrease Accuracy metric (Figure S5), the maximum temperature, minimum temperature, and elevation were the most influential variables. Urban and built-up areas, along with croplands, were the most significant land use types. Overall, the model exhibited a balanced reliance on a diverse set of factors rather than being dominated by one or two key variables, suggesting that multiple covariates contribute significantly to the model\\u0026apos;s predictive accuracy.\\u003c/p\\u003e\\n\\u003ch3\\u003eEnvironmental Suitability\\u003c/h3\\u003e\\n\\u003cp\\u003eWidespread environmental suitability for scrub typhus was predicted across tropical and subtropical regions (Fig. 1). The areas with highest suitability were in Southeast Asia, South Asia, northern Australia and parts of Central Africa. Areas of high suitability (probability\\u0026thinsp;\\u0026gt;\\u0026thinsp;0.5) were also observed across Central and South America, as well as parts of the Caribbean and Northern Australia. Smaller but notable areas of suitability were present in the Middle East, Southern Africa, and portions of East Asia. The traditional understanding of the tsutsugamushi triangle\\u0026mdash;the endemic region for scrub typhus\\u0026mdash;encompasses Southeast Asia, parts of East Asia, northern Australia and the Pacific Islands with a triangular boundary. However, the predicted environmental suitability indicates a broader potential distribution, with significant overlaps and expansions beyond the traditional tsutsugamushi triangle. This suggests that suitable environments for \\u003cem\\u003eOrientia\\u003c/em\\u003e spp. are not confined to this traditional area but also extend beyond it into parts of South Asia, Central and Southern Africa, and South America. This suggests that the risk zones for scrub typhus could be more widespread than previously recognized, underscoring the need for enhanced global surveillance to accurately determine the true extent of the disease\\u0026rsquo;s distribution.\\u003c/p\\u003e\\n\\u003ch3\\u003eReported incidence in high and low suitability areas\\u003c/h3\\u003e\\n\\u003cp\\u003eReported incidence was compared between areas of high and low environmental suitability in Mainland China, South Korea, and Japan. Thailand and Taiwan were excluded from this analysis due to uniformly high population-weighted environmental suitability scores (probability\\u0026thinsp;\\u0026gt;\\u0026thinsp;0.5) across all first-level administrative divisions. The results indicate that regions with environmental suitability probabilities greater than 0.5 (red) consistently exhibited higher mean annual incidences (5.0 per 100,000 population) compared to areas with scores below 0.5 (blue) (0.6 per 100,000 population) (Fig. 2A). This trend was particularly pronounced in Mainland China. A positive relationship was observed, with an increase in environmental suitability associated with a gradual rise in reported incidence as determined by a GLM passion regression analysis conducted across all first-level administrative divisions (Fig. 2B). The shaded area represents the 95% confidence interval, suggesting that while there is some variability in the data, the trend remains statistically significant, especially in higher suitability regions.\\u003c/p\\u003e\\n\\u003cdiv id=\\\"Sec8\\\"\\u003e\\n \\u003ch2\\u003eUncertainty\\u003c/h2\\u003e\\n \\u003cp\\u003eThe absolute uncertainties in the environmental suitability estimates were inherently influenced by the spatial distribution and density of occurrence data, with the highest uncertainties observed in regions with limited data or highly heterogeneous environments, notably in large areas of American, Central and Southern South American, Central Australia, plateaus, rift valleys and deserts in Africa and desert region in the Middle East (Fig. 2 and Figure S6). Furthermore, an analysis of the ratio of the mean to the width of the confidence interval highlighted the primary contributors to relative uncertainty. These were predominantly countries with sparse occurrence points and inconsistent evidence regarding the presence of scrub typhus, such as in remote or data-poor regions.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003ch3\\u003ePopulation living in environmental suitable areas\\u003c/h3\\u003e\\n\\u003cp\\u003eFurther analysis revealed significant variations in environmentally suitable areas and populations living in suitable areas across different WHO regions (Table 1). The full results for 215 countries/regions can be found in Supplementary Information, section 4.3. It was estimated that 4.4 billion people (95% CI: 3.86\\u0026ndash;4.90 billion), about 54% of the global population, live in areas with high environmental suitability for scrub typhus worldwide, while in countries/regions where scrub typhus has been confirmed in humans, 2.5 billion (95% CI: 2.43\\u0026ndash;2.69 billion) people live in environmentally suitable areas. South-East Asia, notably India and Indonesia, has extensive at-risk areas with substantial populations exposed. In the Western Pacific, China, Vietnam and Philippines were major regions of concern, while Brazil dominated in the Americas with the largest area at risk. Some African countries such as Nigeria and Ethiopia also had considerable populations living in high-suitability zones, underscoring the widespread nature of potential exposure across these diverse regions. Most countries in the Americas and Africa have never reported any confirmed human scrub typhus cases (Fig. 4A), but our analysis indicates that large areas in Brazil, the United States, Mexico, Nigeria, Ethiopia, Egypt, DR Congo and Sudan are environmentally suitable for scrub typhus.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eTable 1\\u003c/strong\\u003e Predicted high environmental suitability areas and top 20 highest populations living in environmentally suitable areas by WHO region.\\u003c/p\\u003e\\n\\u003ctable border=\\\"1\\\" cellspacing=\\\"0\\\" cellpadding=\\\"0\\\" width=\\\"591\\\"\\u003e\\n \\u003ctbody\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 92px;\\\"\\u003e\\n \\u003cp\\u003eWHO region\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eCountry\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003eArea (km\\u003csup\\u003e2\\u003c/sup\\u003e)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003ePopulation in millions (uncertainty)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003ePercentage of total population (uncertainty)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eConfirmed human scrub typhus\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd rowspan=\\\"5\\\" style=\\\"width: 92px;\\\"\\u003e\\n \\u003cp\\u003eSouth-East Asia\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eIndia\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e3,231,500\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e1166.32 (1160.26-1169.6)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e98.7% (98.2%-98.9%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eIndonesia\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e1,843,375\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e223.89 (222.02-224.29)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e97.9% (97.0%-98.0%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eBangladesh\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e147,275\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e139.46 (139.19-139.46)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e99.1% (98.9%-99.1%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eThailand\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e531,300\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e57.47 (57.29-57.47)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e98.9% (98.6%-98.9%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eMyanmar (Burma)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e705,525\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e46.14 (45.66-46.15)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e97.5% (96.5%-97.6%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd rowspan=\\\"4\\\" style=\\\"width: 92px;\\\"\\u003e\\n \\u003cp\\u003eWestern Pacific\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eChina\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e1,330,900\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e349.26 (301.13-427.75)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e28.7% (24.7%-35.1%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eVietnam\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e338,050\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e80.33 (79.34-80.38)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e99.2% (98.0%-99.2%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003ePhilippines\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e282,150\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e75.93 (75.75-75.93)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e94.1% (93.9%-94.1%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eSouth Korea\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e109,625\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e40.87 (38.78-41.2)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e96.4% (91.5%-97.2%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd rowspan=\\\"3\\\" style=\\\"width: 92px;\\\"\\u003e\\n \\u003cp\\u003eAmericas\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eBrazil\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e8,466,600\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e175.88 (151.74-177.57)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e96.8% (83.6%-97.8%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eUnited States\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e1,659,100\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e150.06 (82.07-210.98)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e53.5% (29.2%-75.2%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eMexico\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e1,540,975\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e89.77 (63.29-104.9)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e79.5% (56.1%-92.9%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd rowspan=\\\"7\\\" style=\\\"width: 92px;\\\"\\u003e\\n \\u003cp\\u003eAfrica\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eNigeria\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e924,375\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e174.15 (173.68-174.15)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e99.1% (98.8%-99.1%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eEthiopia\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e1,091,200\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e84.51 (64.29-90.98)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e88.2% (67.1%-94.9%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eEgypt\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e527,375\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e83.61 (53.82-86.54)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e96.3% (62.0%-99.7%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eDR Congo\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e2,254,100\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e76.58 (68.53-76.8)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e99.6% (89.1%-99.9%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eTanzania\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e931,150\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e51.77 (46.08-52.25)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e97.7% (87.0%-98.6%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eSudan\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e2,558,600\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e51.2 (47.94-51.23)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e99.9% (93.5%-99.9%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003eKenya\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e579,550\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e42.19 (34.29-42.68)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e97.8% (79.5%-98.9%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd valign=\\\"top\\\" style=\\\"width: 92px;\\\"\\u003e\\n \\u003cp\\u003eEastern Mediterranean\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 78px;\\\"\\u003e\\n \\u003cp\\u003ePakistan\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 85px;\\\"\\u003e\\n \\u003cp\\u003e745,975\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 142px;\\\"\\u003e\\n \\u003cp\\u003e166.44 (148.3-169.2)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 123px;\\\"\\u003e\\n \\u003cp\\u003e93.8% (83.6%-95.4%)\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd valign=\\\"bottom\\\" style=\\\"width: 71px;\\\"\\u003e\\n \\u003cp\\u003eY\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003c/tbody\\u003e\\n\\u003c/table\\u003e\\n\\u003cp\\u003eA list of priority countries was developed where targeted research and enhanced surveillance are most urgently needed to determine if scrub typhus transmission is occurring (Fig. 4A; Supplementary Table S10). With proper surveillance in place, targeted prevention campaigns and treatment guidelines could be implemented to reveal and reduce potential burden associated with scrub typhus. The priority list includes areas where scrub typhus is likely present but underreported, as well as regions with high environmental suitability for transmission but where the disease has not yet been documented (Fig. 4A) and areas which are environmentally suitable but have low accessibility to healthcare facilities (Fig. 4B).\\u003c/p\\u003e\\n\\u003ch3\\u003eEnvironmental suitability changes over time\\u003c/h3\\u003e\\n\\u003cp\\u003eThe global distribution of environmentally suitable areas for scrub typhus has remained largely stable over the past two decades, with consistently widespread regions of high suitability observed across tropical and subtropical zones (Fig. 5A\\u0026ndash;C). Despite some localized expansions and contractions of suitable areas between 2001 and 2020 (Fig. 5D\\u0026ndash;F), the overall extent of environmentally suitable land has remained substantial, with minimal fluctuations in core high-risk regions. Regions in South and Southeast Asia, sub-Saharan Africa, northern Australia, and parts of South America have consistently exhibited high environmental suitability throughout the study period. While minor expansions were observed in areas such as parts of Central Asia and northern South America, these changes did not significantly alter the global distribution pattern. Population exposure has also remained significant, with billions of people residing within environmentally suitable areas over time. Though some shifts in population within suitable areas occurred, the overarching pattern highlights a stable and extensive potential risk zone. Detailed numerical data on changes in suitable area, associated uncertainties, and population exposure across the study period are provided in Supplementary Table S11.\\u003c/p\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eThis study presents the first high-resolution global environmental suitability map for scrub typhus, revealing that substantial areas of high suitability extend beyond the traditionally recognized endemic regions. Previous research has largely focused on the Asia-Pacific region\\u003csup\\u003e\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR30\\\" class=\\\"CitationRef\\\"\\u003e30\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR52\\\" class=\\\"CitationRef\\\"\\u003e52\\u003c/span\\u003e\\u003c/sup\\u003e with only a few descriptive studies addressing the global epidemiology of scrub typhus\\u003csup\\u003e\\u003cspan citationid=\\\"CR53\\\" class=\\\"CitationRef\\\"\\u003e53\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR54\\\" class=\\\"CitationRef\\\"\\u003e54\\u003c/span\\u003e\\u003c/sup\\u003e. This study build on these efforts by expanding the geographic scope globally. By integrating a comprehensive database of scrub typhus occurrences with an extensive array of environmental covariates, advanced machine learning models were employed to predict areas of high environmental suitability for scrub typhus occurrence. The results underscore the need for a re-evaluation of the global distribution of this neglected tropical disease, which has implications for public health strategies and disease burden estimation.\\u003c/p\\u003e \\u003cp\\u003eOne of the key findings from this analysis is the extensive suitability of environments for scrub typhus across regions that have not historically reported cases. Notably, large areas in Brazil, the United States, Mexico, Nigeria, Ethiopia, and Egypt\\u0026mdash;countries with no documented local transmission\\u0026mdash;were identified as potential high-risk zones suitable for scrub typhus occurrence. These findings challenge the traditional confines of the \\\"tsutsugamushi triangle\\\" and highlight the potential for scrub typhus to emerge in regions previously considered non-endemic. The expansion of scrub typhus beyond the traditional endemic regions, as evidenced by the recent identification of new \\u003cem\\u003eOrientia\\u003c/em\\u003e species in Dubai\\u003csup\\u003e\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e\\u003c/sup\\u003e and Chile\\u003csup\\u003e\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e, the detection of \\u003cem\\u003eOrientia\\u003c/em\\u003e in field collected free-living \\u003cem\\u003eEutrombicula\\u003c/em\\u003e chiggers in the United States\\u003csup\\u003e\\u003cspan citationid=\\\"CR55\\\" class=\\\"CitationRef\\\"\\u003e55\\u003c/span\\u003e\\u003c/sup\\u003e and emerging serologic and molecular evidence of \\u003cem\\u003eOrientia\\u003c/em\\u003e spp. endemicity in Uganda\\u003csup\\u003e\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e\\u003c/sup\\u003e supports the hypothesis that scrub typhus is not confined to the \\\"tsutsugamushi triangle\\\".\\u003c/p\\u003e \\u003cp\\u003eDespite the identification of large environmentally suitable areas for scrub typhus transmission in the Americas and Africa, the number of reported human cases in many countries in these regions remains low or zero. Potential explanations include limited healthcare access, competing national health priorities, historical under-recognition of scrub typhus and the possible absence or low density of suitable reservoir hosts or vector arthropods (e.g., chigger mites) that occasionally bite humans. In the United States, the issue is unlikely to be related to healthcare access or national healthcare expenditure, as most of the population has access to advanced diagnostic facilities. While under-recognition remains a possible factor- given scrub typhus\\u0026rsquo; nonspecific febrile symptoms, which can easily be mistaken for more common tick-borne diseases such as Lyme disease or spotted fever group rickettsiosis, which are endemic to parts of the U.S\\u003csup\\u003e\\u003cspan citationid=\\\"CR56\\\" class=\\\"CitationRef\\\"\\u003e56\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR57\\\" class=\\\"CitationRef\\\"\\u003e57\\u003c/span\\u003e\\u003c/sup\\u003e. Additionally, the historical focus of scrub typhus research has been centred on the Asia-Pacific region and it was long assumed that \\u003cem\\u003eO. tsutsugamushi\\u003c/em\\u003e was restricted to this region. As a result, diagnostic protocols in the U.S. may not routinely include tests for \\u003cem\\u003eOrientia\\u003c/em\\u003e species unless there is travel history to endemic areas\\u003csup\\u003e\\u003cspan citationid=\\\"CR58\\\" class=\\\"CitationRef\\\"\\u003e58\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR59\\\" class=\\\"CitationRef\\\"\\u003e59\\u003c/span\\u003e\\u003c/sup\\u003e. Alternative explanations may include the presence of mild or asymptomatic \\u003cem\\u003eOrientia\\u003c/em\\u003e infections that do not prompt healthcare-seeking behaviour or the existence of local vector species that are unsuitable for efficient human transmission.\\u003c/p\\u003e \\u003cp\\u003eIn contrast, in Africa, healthcare access and national health priorities may play a larger role in the underreporting of scrub typhus. Many regions in Africa face severe challenges in providing even basic healthcare services, with long travel times to healthcare facilities and chronic underfunding of public health systems\\u003csup\\u003e\\u003cspan citationid=\\\"CR60\\\" class=\\\"CitationRef\\\"\\u003e60\\u003c/span\\u003e\\u003c/sup\\u003e. Diseases like Ebola, malaria, HIV/AIDS, and tuberculosis, which have higher mortality rates and more severe public health impacts, take priority in resource allocation and healthcare response\\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR62 CR63\\\" citationid=\\\"CR61\\\" class=\\\"CitationRef\\\"\\u003e61\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR64\\\" class=\\\"CitationRef\\\"\\u003e64\\u003c/span\\u003e\\u003c/sup\\u003e. Moreover, the historical relative neglect of scrub typhus scientific research in Africa and the lack of diagnostic tools and awareness among healthcare providers further exacerbates this issue. It is thus possible that scrub typhus is being misdiagnosed and underreported, even though the environmental conditions may support the transmission of \\u003cem\\u003eOrientia\\u003c/em\\u003e.\\u003c/p\\u003e \\u003cp\\u003eThe potential role of climate change and environmental conditions in driving this expanded range cannot be overlooked. Changes in temperature, precipitation patterns, and habitat availability due to climate change may create new ecological niches for \\u003cem\\u003eOrientia\\u003c/em\\u003e and its vectors, facilitating the spread of the disease to previously unaffected areas\\u003csup\\u003e\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e\\u003c/sup\\u003e. In addition, migratory birds may play a critical role in spreading infected mites or vectors across wide geographical areas\\u003csup\\u003e\\u003cspan citationid=\\\"CR65\\\" class=\\\"CitationRef\\\"\\u003e65\\u003c/span\\u003e\\u003c/sup\\u003e and contribute to the creation of new hotspots for scrub typhus, particularly in regions where suitable environmental conditions are emerging due to climate change\\u003csup\\u003e\\u003cspan citationid=\\\"CR66\\\" class=\\\"CitationRef\\\"\\u003e66\\u003c/span\\u003e\\u003c/sup\\u003e. The relative importance of different covariates in the model, reveals key insights into the factors that most strongly influence the predictive accuracy of scrub typhus environmental suitability. Temperature variables, both maximum temperature and minimum temperature, were shown to have the highest impact on model accuracy, underscoring the critical role of climate in shaping the disease\\u0026rsquo;s distribution. This is consistent with the biology of \\u003cem\\u003eO. tsutsugamushi\\u003c/em\\u003e and its vectors: warmer and more humid climates tend to favour mite population growth by providing optimal conditions for egg-laying and larvae survival\\u003csup\\u003e\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e,\\u003cspan additionalcitationids=\\\"CR68\\\" citationid=\\\"CR67\\\" class=\\\"CitationRef\\\"\\u003e67\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR69\\\" class=\\\"CitationRef\\\"\\u003e69\\u003c/span\\u003e\\u003c/sup\\u003e. Elevation also ranks highly, as higher elevations tend to be inhospitable for both vector and host survival, with fewer people and lower vegetation cover, creating a natural barrier to scrub typhus transmission\\u003csup\\u003e\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR31\\\" class=\\\"CitationRef\\\"\\u003e31\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR70\\\" class=\\\"CitationRef\\\"\\u003e70\\u003c/span\\u003e\\u003c/sup\\u003e. Socioeconomic factors such as urban accessibility and population density, emerge as significant contributors by elevating the chance of potential human-vector interactions\\u003csup\\u003e\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR31\\\" class=\\\"CitationRef\\\"\\u003e31\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR71\\\" class=\\\"CitationRef\\\"\\u003e71\\u003c/span\\u003e\\u003c/sup\\u003e. Vegetation density, such as Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI), along with vegetation types, such as evergreen broadleaf forests and croplands, contribute to the ecological niches where both the vector and the pathogen may persist\\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR73\\\" citationid=\\\"CR72\\\" class=\\\"CitationRef\\\"\\u003e72\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR74\\\" class=\\\"CitationRef\\\"\\u003e74\\u003c/span\\u003e\\u003c/sup\\u003e. Areas with minimal vegetation, like barren lands, seem to have the lowest relative importance, possibly due to lower vector presence or limited human habitation. These associations have been previously recognized in localized studies, but our findings demonstrate their relevance at a global scale, highlighting the importance of incorporating a wide range of covariates when assessing scrub typhus risk.\\u003c/p\\u003e \\u003cp\\u003eThe uncertainties in our environmental suitability estimates, especially in regions with sparse human case occurrence data, underscore the difficulties in accurately and confidently predicting the disease suitability distribution. The widest confidence intervals were observed in areas where there is inconsistent or limited evidence of scrub typhus presence. This was most pronounced in regions where environmental suitability was consistently high across models, yet lacked robust occurrence records to validate predictions. For example, much of Africa, parts of South America, and central Asia exhibited high environmental suitability but suffer from a lack of confirmed cases, leading to increased uncertainty in these areas. Improving data collection efforts in these regions could greatly refine our understanding of the spatial distribution of scrub typhus and, consequently, improve the precision of disease burden estimates. There are several key areas on which these efforts should focus. First, targeted research is needed to confirm the presence of scrub typhus in regions where environmental suitability is high, but occurrence data is lacking. This includes field studies and epidemiological investigations to uncover undetected cases\\u003csup\\u003e\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e. Second, increasing awareness among physicians and communities in endemic but likely underreported regions is essential for improving diagnosis and reporting\\u003csup\\u003e\\u003cspan citationid=\\\"CR75\\\" class=\\\"CitationRef\\\"\\u003e75\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR76\\\" class=\\\"CitationRef\\\"\\u003e76\\u003c/span\\u003e\\u003c/sup\\u003e. Enhanced training on recognizing scrub typhus symptoms and the availability of appropriate diagnostic testing, particularly in areas with limited healthcare infrastructure, could significantly reduce underreporting\\u003csup\\u003e\\u003cspan citationid=\\\"CR77\\\" class=\\\"CitationRef\\\"\\u003e77\\u003c/span\\u003e\\u003c/sup\\u003e. Engaging communities in health education and sharing research knowledge can further strengthen these efforts, improving overall awareness and encouraging timely healthcare-seeking behaviour\\u003csup\\u003e\\u003cspan citationid=\\\"CR78\\\" class=\\\"CitationRef\\\"\\u003e78\\u003c/span\\u003e\\u003c/sup\\u003e. Finally, strengthening national and regional surveillance systems will be crucial for systematically capturing both confirmed and suspected cases. Investments in expanding testing availability and improving reporting mechanisms can lead to a more accurate picture of the disease's global distribution, ultimately enhancing the precision of future models and supporting effective public health interventions.\\u003c/p\\u003e \\u003cp\\u003eMoreover, this analysis revealed that approximately 2.5\\u0026nbsp;billion people (95% CI: 2.43\\u0026ndash;2.69\\u0026nbsp;billion) are currently living in environmentally suitable areas within countries or regions where human cases of scrub typhus have already been confirmed. This figure provides the first data-driven estimate of land area and population at risk for scrub typhus, in contrast to the only previously available generalized statement from a 1997 study, which stated that \\\"about 13\\u0026nbsp;million square kilometres of land are endemic, and more than a billion people would appear to be at risk\\\" within the tsutsugamushi triangle\\u003csup\\u003e\\u003cspan citationid=\\\"CR79\\\" class=\\\"CitationRef\\\"\\u003e79\\u003c/span\\u003e\\u003c/sup\\u003e. This new estimate incorporates not only the traditional areas within the triangle but also countries and regions outside of it where human cases of scrub typhus have been confirmed, or suspected, in recent years. Given the 27 years that have passed since the original estimate, during which global populations have risen and climate has changed substantially, particularly in many endemic countries\\u003csup\\u003e\\u003cspan citationid=\\\"CR80\\\" class=\\\"CitationRef\\\"\\u003e80\\u003c/span\\u003e\\u003c/sup\\u003e, this updated figure of 2.5\\u0026nbsp;billion people highlights the growing significance of scrub typhus as a public health threat. Expanding the scope of the analysis globally, it was estimated that over 4.4\\u0026nbsp;billion people live in high environmental suitability areas. While this large number highlights the widespread nature of suitable environments for the disease, it does not imply that large numbers of people are at high risk of infection. Contracting scrub typhus requires several other factors, including exposure to a suitable vector (infected and human-biting), the presence of the disease in the environment, and other ecological and human-mediated pathways\\u003csup\\u003e\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e\\u003c/sup\\u003e. Therefore, while the environmental suitability is high, quantifying the actual risk of infection requires consideration of a complex interplay of additional factors.\\u003c/p\\u003e \\u003cp\\u003eDespite these extensive efforts to compile comprehensive data on scrub typhus occurrence and employ new modelling approaches to enhance the predictive power of these data, several limitations remain. The empirical evidence base for risk remains limited by the availability and quality of georeferenced occurrence data, which vary significantly across regions. Notably, a substantial portion of the data originate from Asia while significant gaps remain in remote or underreported areas, which may introduce bias when extrapolating findings to other regions with different ecological and epidemiological contexts. The developed model is likely more accurate in well-surveyed/data-rich regions like Southeast Asia, but less reliable in data-sparse areas such as Africa and the Americas. Additionally, while a broad range of global environmental covariates was utilized, they may not capture local variations that could significantly influence disease transmission dynamics. The regional variability in vector and host ecology was not accounted for, which may limit the model's applicability in predicting actual disease risk. Furthermore, underdiagnosis of asymptomatic or mild cases, particularly in areas with limited surveillance, could skew the estimates. These limitations emphasize the need for improved surveillance in underrepresented regions and ongoing research to refine the environmental and biological factors driving scrub typhus transmission, especially in the context of climate change. While this study provides the most comprehensive, evidence-based estimate based on currently available data and methodologies, continued advancements in surveillance, diagnostics, and data collection, alongside the development of new analytical tools, will be essential for refining and improving future assessments.\\u003c/p\\u003e \\u003cp\\u003eIn conclusion, this study provides updated and more in-depth understanding of the global distribution of scrub typhus and highlights the need for more comprehensive surveillance and further investigations. The global burden of scrub typhus could be heavily underestimated and represents a growing challenge to public health officials and policymakers. The high-resolution environmental suitability maps developed in this study offer valuable insights to prioritize regions for further investigation, strengthened surveillance and intervention and to better estimate the global burden of this neglected disease. Success in tackling this growing global threat is, in part, contingent on strengthening the evidence base on which control planning decisions and their impact are evaluated. It is hoped that this evaluation of contemporary environmental suitability will help to advance that goal.\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eAcknowledgements\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThis research was funded in part by the Wellcome Trust [220211],\\u0026nbsp;National Natural Science Foundation of China [42201497] and Youth Innovation Promotion Association [2023000117]. The funder had no role in study design, data collection, data analysis, data interpretation, or writing of this study. For the purpose of Open Access, the author has applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission. We would like to thank Thailand Ministry of Public Health for providing the national surveillance data. All authors read and approved the final manuscript.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eContributions\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eRM, BS, and ND conceived the study and provided overall guidance. QW, TM, FD and CZ contributed to data collection. QW had major roles in formulating the analysis under the supervision of RM, BS, and ND. QW prepared the first draft and finalized the manuscript based on feedback from all co-authors. All co-authors have contributed substantially to the review and editing of the manuscript.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eEthics declarations\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCompeting interests\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe declare no competing interests.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eData availability\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe authors had full access to the data and were responsible for conducting the analyses. This research was conducted in compliance with all relevant ethical regulations. Ethical approval is not applicable. These de-identified data may be made available upon reasonable request via a proposal-based process. Interested researchers can contact qian@tropmedres.ac.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n\\u003cli\\u003eXu, G., Walker, D.H., Jupiter, D., Melby, P.C. \\u0026amp; Arcari, C.M. A review of the global epidemiology of scrub typhus. \\u003cem\\u003ePLoS Neglected Tropical Diseases\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e(2017).\\u003c/li\\u003e\\n\\u003cli\\u003eElliott, I.\\u003cem\\u003e, et al.\\u003c/em\\u003e Scrub typhus ecology: a systematic review of Orientia in vectors and hosts. \\u003cem\\u003eParasites \\u0026amp; vectors\\u003c/em\\u003e \\u003cstrong\\u003e12\\u003c/strong\\u003e, 513 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eRajapakse, S., Weeratunga, P., Sivayoganathan, S. \\u0026amp; Fernando, S.D. Clinical manifestations of scrub typhus. \\u003cem\\u003eTransactions of the Royal Society of Tropical Medicine and Hygiene\\u003c/em\\u003e \\u003cstrong\\u003e111\\u003c/strong\\u003e, 43-54 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eKelly, D.J., Fuerst, P.A., Ching, W.-M. \\u0026amp; Richards, A.L. Scrub typhus: the geographic distribution of phenotypic and genotypic variants of Orientia tsutsugamushi. \\u003cem\\u003eClinical infectious diseases\\u003c/em\\u003e \\u003cstrong\\u003e48\\u003c/strong\\u003e, S203-S230 (2009).\\u003c/li\\u003e\\n\\u003cli\\u003eFan, M.Y., Walker, D.H., Yu, S.R. \\u0026amp; Liu, Q.H. Epidemiology and ecology of rickettsial diseases in the People\\u0026apos;s Republic of China. \\u003cem\\u003eReviews of infectious diseases\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, 823-840 (1987).\\u003c/li\\u003e\\n\\u003cli\\u003ePhilip, C.B. Observations on Tsutsugamushi Disease (Mite-borne or Scrub Typhus) in northwest Honshu Island, Japan, in the Fall of 1945. I. Epidemiological and ecological Data. \\u003cem\\u003eAmerican Journal of Hygiene\\u003c/em\\u003e \\u003cstrong\\u003e46\\u003c/strong\\u003e, 45-pp (1947).\\u003c/li\\u003e\\n\\u003cli\\u003eWangrangsimakul, T.\\u003cem\\u003e, et al.\\u003c/em\\u003e The estimated burden of scrub typhus in Thailand from national surveillance data (2003-2018). \\u003cem\\u003ePLoS Neglected Tropical Diseases\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e, 1-20 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eDevasagayam, E.\\u003cem\\u003e, et al.\\u003c/em\\u003e The burden of scrub typhus in India: A systematic review. \\u003cem\\u003ePLoS neglected tropical diseases\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, e0009619 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eIzzard, L.\\u003cem\\u003e, et al.\\u003c/em\\u003e Isolation of a novel Orientia species (O. chuto sp. nov.) from a patient infected in Dubai. \\u003cem\\u003eJ Clin Microbiol\\u003c/em\\u003e \\u003cstrong\\u003e48\\u003c/strong\\u003e, 4404-4409 (2010).\\u003c/li\\u003e\\n\\u003cli\\u003eWeitzel, T.\\u003cem\\u003e, et al.\\u003c/em\\u003e Endemic Scrub Typhus in South America. \\u003cem\\u003eThe New England journal of medicine\\u003c/em\\u003e \\u003cstrong\\u003e375\\u003c/strong\\u003e, 954-961 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eBlair, P.W.\\u003cem\\u003e, et al.\\u003c/em\\u003e Evidence of Orientia spp. Endemicity among Severe Infectious Disease Cohorts, Uganda. \\u003cem\\u003eEmerg Infect Dis\\u003c/em\\u003e \\u003cstrong\\u003e30\\u003c/strong\\u003e, 1442-1446 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eIzzard, L.\\u003cem\\u003e, et al.\\u003c/em\\u003e Isolation of a novel Orientia species (O. chuto sp. nov.) from a patient infected in Dubai. \\u003cem\\u003eJournal of clinical microbiology\\u003c/em\\u003e \\u003cstrong\\u003e48\\u003c/strong\\u003e, 4404-4409 (2010).\\u003c/li\\u003e\\n\\u003cli\\u003eAbarca, K.\\u003cem\\u003e, et al.\\u003c/em\\u003e Molecular Description of a Novel Orientia Species Causing Scrub Typhus in Chile. \\u003cem\\u003eEmerg Infect Dis\\u003c/em\\u003e \\u003cstrong\\u003e26\\u003c/strong\\u003e, 2148-2156 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eWangrangsimakul, T.\\u003cem\\u003e, et al.\\u003c/em\\u003e The estimated burden of scrub typhus in Thailand from national surveillance data(2003-2018). \\u003cem\\u003ePLoS neglected tropical diseases\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e, e0008233 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eLi, Z.\\u003cem\\u003e, et al.\\u003c/em\\u003e Epidemiologic Changes of Scrub Typhus in China, 1952-2016. \\u003cem\\u003eEmerging infectious diseases\\u003c/em\\u003e \\u003cstrong\\u003e26\\u003c/strong\\u003e, 1091-1101 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eYu, H.\\u003cem\\u003e, et al.\\u003c/em\\u003e Scrub typhus in Jiangsu Province, China: epidemiologic features and spatial risk analysis. \\u003cem\\u003eBMC Infectious Diseases\\u003c/em\\u003e \\u003cstrong\\u003e18\\u003c/strong\\u003e, 372 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eAcharya, B.K.\\u003cem\\u003e, et al.\\u003c/em\\u003e Mapping Environmental Suitability of Scrub Typhus in Nepal Using MaxEnt and Random Forest Models. \\u003cem\\u003eInternational journal of environmental research and public health\\u003c/em\\u003e \\u003cstrong\\u003e16\\u003c/strong\\u003e(2019).\\u003c/li\\u003e\\n\\u003cli\\u003eXin, H.\\u003cem\\u003e, et al.\\u003c/em\\u003e Risk mapping of scrub typhus infections in Qingdao city, China. \\u003cem\\u003ePLoS neglected tropical diseases\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e, e0008757 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eWang, Q.\\u003cem\\u003e, et al.\\u003c/em\\u003e A systematic review of environmental covariates and methods for spatial or temporal scrub typhus distribution prediction. \\u003cem\\u003eEnviron Res\\u003c/em\\u003e \\u003cstrong\\u003e263\\u003c/strong\\u003e, 120067 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eDing, F.\\u003cem\\u003e, et al.\\u003c/em\\u003e Climate drives the spatiotemporal dynamics of scrub typhus in China. \\u003cem\\u003eGlob Chang Biol\\u003c/em\\u003e \\u003cstrong\\u003e28\\u003c/strong\\u003e, 6618-6628 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eOgawa, T.\\u003cem\\u003e, et al.\\u003c/em\\u003e Analysis of Differences in Characteristics of High-Risk Endemic Areas for Contracting Japanese Spotted Fever, Tsutsugamushi Disease, and Severe Fever With Thrombocytopenia Syndrome. \\u003cem\\u003eOpen Forum Infect Dis\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, ofae025 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eRoberts, T.\\u003cem\\u003e, et al.\\u003c/em\\u003e A spatio-temporal analysis of scrub typhus and murine typhus in Laos; implications from changing landscapes and climate. \\u003cem\\u003ePLoS neglected tropical diseases\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, e0009685 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eJ, Q.\\u003cem\\u003e, et al.\\u003c/em\\u003e Spatiotemporal heterogeneity and long-term impact of meteorological, environmental, and socio-economic factors on scrub typhus in China from 2006 to 2018. \\u003cem\\u003eBMC public health\\u003c/em\\u003e \\u003cstrong\\u003e24\\u003c/strong\\u003e, 538 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eChang, T., Min, K.D., Cho, S.I. \\u0026amp; Kim, Y. Associations of meteorological factors and dynamics of scrub typhus incidence in South Korea: A nationwide time-series study. \\u003cem\\u003eEnviron Res\\u003c/em\\u003e \\u003cstrong\\u003e245\\u003c/strong\\u003e, 117994 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eMungmungpuntipantip, R. \\u0026amp; Wiwanitkit, V. Correlation between rainfall and the prevalence of scrub typhus: an observation from a tropical endemic country. \\u003cem\\u003eInt.j.med.surg.sci.(Print)\\u003c/em\\u003e \\u003cstrong\\u003e8\\u003c/strong\\u003e, 1-4 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003e吴义城, 张文义 \\u0026amp; 李申龙. 中国人民解放军军事医学科学院 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eWei, Y.\\u003cem\\u003e, et al.\\u003c/em\\u003e Climate variability, animal reservoir and transmission of scrub typhus in Southern China. \\u003cem\\u003ePLoS Neglected Tropical Diseases\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, e0005447 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eLu, J., Liu, Y., Ma, X., Li, M. \\u0026amp; Yang, Z. Impact of Meteorological Factors and Southern Oscillation Index on Scrub Typhus Incidence in Guangzhou, Southern China, 2006-2018. \\u003cem\\u003eFront Med (Lausanne)\\u003c/em\\u003e \\u003cstrong\\u003e8\\u003c/strong\\u003e, 667549 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eMin, K.-D., Lee, J.-Y., So, Y. \\u0026amp; Cho, S.-I. Deforestation Increases the Risk of Scrub Typhus in Korea. \\u003cem\\u003eInternational journal of environmental research and public health\\u003c/em\\u003e \\u003cstrong\\u003e16\\u003c/strong\\u003e(2019).\\u003c/li\\u003e\\n\\u003cli\\u003eWangrangsimakul, T.\\u003cem\\u003e, et al.\\u003c/em\\u003e The estimated burden of scrub typhus in Thailand from national surveillance data (2003-2018). \\u003cem\\u003ePLoS Negl Trop Dis\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e, e0008233 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eZheng, C., Jiang, D., Ding, F., Fu, J. \\u0026amp; Hao, M. Spatiotemporal Patterns and Risk Factors for Scrub Typhus From 2007 to 2017 in Southern China. \\u003cem\\u003eClinical infectious diseases : an official publication of the Infectious Diseases Society of America\\u003c/em\\u003e \\u003cstrong\\u003e69\\u003c/strong\\u003e, 1205-1211 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003e刘晓宁. 硕士, 安徽医科大学 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eHuang, X.\\u003cem\\u003e, et al.\\u003c/em\\u003e Prediction of risk factors for scrub typhus from 2006 to 2019 based on random forest model in Guangzhou, China. \\u003cem\\u003eTropical Medicine \\u0026amp; International Health\\u003c/em\\u003e \\u003cstrong\\u003e28\\u003c/strong\\u003e, 551-561 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003eStevens, G.A.\\u003cem\\u003e, et al.\\u003c/em\\u003e Guidelines for Accurate and Transparent Health Estimates Reporting: the GATHER statement. \\u003cem\\u003eThe Lancet\\u003c/em\\u003e \\u003cstrong\\u003e388\\u003c/strong\\u003e, e19-e23 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eWang, Q.\\u003cem\\u003e, et al.\\u003c/em\\u003e Global and regional seroprevalence, incidence, mortality of, and risk factors for scrub typhus: A systematic review and meta-analysis. \\u003cem\\u003eInt J Infect Dis\\u003c/em\\u003e \\u003cstrong\\u003e146\\u003c/strong\\u003e, 107151 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eAra\\u0026uacute;jo, M.B. \\u0026amp; New, M. Ensemble forecasting of species distributions. \\u003cem\\u003eTrends Ecol Evol\\u003c/em\\u003e \\u003cstrong\\u003e22\\u003c/strong\\u003e, 42-47 (2007).\\u003c/li\\u003e\\n\\u003cli\\u003eHastie, T.J. Generalized additive models. in \\u003cem\\u003eStatistical models in S\\u003c/em\\u003e 249-307 (Routledge, 2017).\\u003c/li\\u003e\\n\\u003cli\\u003eElith, J., Leathwick, J.R. \\u0026amp; Hastie, T. A working guide to boosted regression trees. \\u003cem\\u003eJournal of animal ecology\\u003c/em\\u003e \\u003cstrong\\u003e77\\u003c/strong\\u003e, 802-813 (2008).\\u003c/li\\u003e\\n\\u003cli\\u003eBreiman, L. Random forests. \\u003cem\\u003eMachine learning\\u003c/em\\u003e \\u003cstrong\\u003e45\\u003c/strong\\u003e, 5-32 (2001).\\u003c/li\\u003e\\n\\u003cli\\u003ePrasad, A.M., Iverson, L.R. \\u0026amp; Liaw, A. Newer classification and regression tree techniques: bagging and random forests for ecological prediction. \\u003cem\\u003eEcosystems\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, 181-199 (2006).\\u003c/li\\u003e\\n\\u003cli\\u003eTibshirani, R. Regression shrinkage and selection via the lasso. \\u003cem\\u003eJournal of the Royal Statistical Society Series B: Statistical Methodology\\u003c/em\\u003e \\u003cstrong\\u003e58\\u003c/strong\\u003e, 267-288 (1996).\\u003c/li\\u003e\\n\\u003cli\\u003eNelder, J.A. \\u0026amp; Wedderburn, R.W. Generalized linear models. \\u003cem\\u003eJournal of the Royal Statistical Society Series A: Statistics in Society\\u003c/em\\u003e \\u003cstrong\\u003e135\\u003c/strong\\u003e, 370-384 (1972).\\u003c/li\\u003e\\n\\u003cli\\u003eHastie, T., Tibshirani, R., Friedman, J.H. \\u0026amp; Friedman, J.H. \\u003cem\\u003eThe elements of statistical learning: data mining, inference, and prediction\\u003c/em\\u003e, (Springer, 2009).\\u003c/li\\u003e\\n\\u003cli\\u003eBhatt, S.\\u003cem\\u003e, et al.\\u003c/em\\u003e Improved prediction accuracy for disease risk mapping using Gaussian process stacked generalization. \\u003cem\\u003eJ R Soc Interface\\u003c/em\\u003e \\u003cstrong\\u003e14\\u003c/strong\\u003e(2017).\\u003c/li\\u003e\\n\\u003cli\\u003eZhou, Z.-H. \\u003cem\\u003eEnsemble methods: foundations and algorithms\\u003c/em\\u003e, (CRC press, 2025).\\u003c/li\\u003e\\n\\u003cli\\u003eWolpert, D.H. Stacked generalization. \\u003cem\\u003eNeural networks\\u003c/em\\u003e \\u003cstrong\\u003e5\\u003c/strong\\u003e, 241-259 (1992).\\u003c/li\\u003e\\n\\u003cli\\u003eWilson, E.B. Probable Inference, the Law of Succession, and Statistical Inference. \\u003cem\\u003eJournal of the American Statistical Association\\u003c/em\\u003e \\u003cstrong\\u003e22\\u003c/strong\\u003e, 209-212 (1927).\\u003c/li\\u003e\\n\\u003cli\\u003eEfron, B. \\u0026amp; Tibshirani, R.J. \\u003cem\\u003eAn introduction to the bootstrap\\u003c/em\\u003e, (Chapman and Hall/CRC, 1994).\\u003c/li\\u003e\\n\\u003cli\\u003ePhillips, S.J.\\u003cem\\u003e, et al.\\u003c/em\\u003e Sample selection bias and presence‐only distribution models: implications for background and pseudo‐absence data. \\u003cem\\u003eEcological applications\\u003c/em\\u003e \\u003cstrong\\u003e19\\u003c/strong\\u003e, 181-197 (2009).\\u003c/li\\u003e\\n\\u003cli\\u003eWilliams, J.N.\\u003cem\\u003e, et al.\\u003c/em\\u003e Using species distribution models to predict new occurrences for rare plants. \\u003cem\\u003eDiversity and distributions\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, 565-576 (2009).\\u003c/li\\u003e\\n\\u003cli\\u003eWeiss, D.J.\\u003cem\\u003e, et al.\\u003c/em\\u003e Global maps of travel time to healthcare facilities. \\u003cem\\u003eNature Medicine\\u003c/em\\u003e \\u003cstrong\\u003e26\\u003c/strong\\u003e, 1835-1838 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eDevasagayam, E.\\u003cem\\u003e, et al.\\u003c/em\\u003e The burden of scrub typhus in India: A systematic review. \\u003cem\\u003ePLoS Negl Trop Dis\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, e0009619 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eBonell, A., Lubell, Y., Newton, P.N., Crump, J.A. \\u0026amp; Paris, D.H. Estimating the burden of scrub typhus: A systematic review. \\u003cem\\u003ePLoS Negl Trop Dis\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, e0005838 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eXu, G., Walker, D.H., Jupiter, D., Melby, P.C. \\u0026amp; Arcari, C.M. A review of the global epidemiology of scrub typhus. \\u003cem\\u003ePLoS Negl Trop Dis\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, e0006062 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eChen, K.\\u003cem\\u003e, et al.\\u003c/em\\u003e Detection of Orientia spp. Bacteria in Field-Collected Free-Living Eutrombicula Chigger Mites, United States. \\u003cem\\u003eEmerg Infect Dis\\u003c/em\\u003e \\u003cstrong\\u003e29\\u003c/strong\\u003e, 1676-1679 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003eSchwartz, A.M. Surveillance for lyme disease\\u0026mdash;United States, 2008\\u0026ndash;2015. \\u003cem\\u003eMMWR. Surveillance Summaries\\u003c/em\\u003e \\u003cstrong\\u003e66\\u003c/strong\\u003e(2017).\\u003c/li\\u003e\\n\\u003cli\\u003eBiggs, H.M. Diagnosis and management of tickborne rickettsial diseases: Rocky Mountain spotted fever and other spotted fever group rickettsioses, ehrlichioses, and anaplasmosis\\u0026mdash;United States. \\u003cem\\u003eMMWR. Recommendations and Reports\\u003c/em\\u003e \\u003cstrong\\u003e65\\u003c/strong\\u003e(2016).\\u003c/li\\u003e\\n\\u003cli\\u003eHendershot, E.F. \\u0026amp; Sexton, D.J. Scrub typhus and rickettsial diseases in international travelers: a review. \\u003cem\\u003eCurrent infectious disease reports\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, 66-72 (2009).\\u003c/li\\u003e\\n\\u003cli\\u003eSultana, R.\\u003cem\\u003e, et al.\\u003c/em\\u003e The Brief Case: A traveler\\u0026rsquo;s tale\\u0026mdash;imported scrub typhus in a child returning from Bangladesh. \\u003cem\\u003eJournal of Clinical Microbiology\\u003c/em\\u003e \\u003cstrong\\u003e61\\u003c/strong\\u003e, e00359-00323 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003eWeiss, D.\\u003cem\\u003e, et al.\\u003c/em\\u003e Global maps of travel time to healthcare facilities. \\u003cem\\u003eNature medicine\\u003c/em\\u003e \\u003cstrong\\u003e26\\u003c/strong\\u003e, 1835-1838 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eTeam, W.E.R. Ebola virus disease in West Africa\\u0026mdash;the first 9 months of the epidemic and forward projections. \\u003cem\\u003eNew England Journal of Medicine\\u003c/em\\u003e \\u003cstrong\\u003e371\\u003c/strong\\u003e, 1481-1495 (2014).\\u003c/li\\u003e\\n\\u003cli\\u003eBhatt, S.\\u003cem\\u003e, et al.\\u003c/em\\u003e The effect of malaria control on Plasmodium falciparum in Africa between 2000 and 2015. \\u003cem\\u003eNature\\u003c/em\\u003e \\u003cstrong\\u003e526\\u003c/strong\\u003e, 207-211 (2015).\\u003c/li\\u003e\\n\\u003cli\\u003eHotez, P.J. \\u0026amp; Kamath, A. Neglected tropical diseases in sub-Saharan Africa: review of their prevalence, distribution, and disease burden. \\u003cem\\u003ePLoS neglected tropical diseases\\u003c/em\\u003e \\u003cstrong\\u003e3\\u003c/strong\\u003e, e412 (2009).\\u003c/li\\u003e\\n\\u003cli\\u003eGandhi, N.R.\\u003cem\\u003e, et al.\\u003c/em\\u003e Extensively drug-resistant tuberculosis as a cause of death in patients co-infected with tuberculosis and HIV in a rural area of South Africa. \\u003cem\\u003eThe Lancet\\u003c/em\\u003e \\u003cstrong\\u003e368\\u003c/strong\\u003e, 1575-1580 (2006).\\u003c/li\\u003e\\n\\u003cli\\u003eScott, J.D. Studies abound on how far north Ixodes scapularis ticks are transported by birds. \\u003cem\\u003eTicks and tick-borne diseases\\u003c/em\\u003e \\u003cstrong\\u003e7\\u003c/strong\\u003e, 327-328 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eWalker, D.H. Scrub Typhus \\u0026mdash; Scientific Neglect, Ever-Widening Impact. \\u003cem\\u003eNew England Journal of Medicine\\u003c/em\\u003e \\u003cstrong\\u003e375\\u003c/strong\\u003e, 913-915 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eAudy, J.R. The ecology of scrub typhus. \\u003cem\\u003eStudies in disease ecology\\u003c/em\\u003e \\u003cstrong\\u003e2\\u003c/strong\\u003e, 389-432 (1961).\\u003c/li\\u003e\\n\\u003cli\\u003eWharton, G.W. \\u0026amp; Fuller, H.S. A Manual of the Chiggers. The Biology, Classification, Distribution, and Importance to Man of the Larvae of the Family Trombiculidae (Acari\\u0026ntilde;a). (1952).\\u003c/li\\u003e\\n\\u003cli\\u003eKawamura, R. \\u0026amp; Ikeda, K. Ecological Study of the Tsutsugamushi, Trombicula akamushi (Brumpt). (1936).\\u003c/li\\u003e\\n\\u003cli\\u003eL, L.\\u003cem\\u003e, et al.\\u003c/em\\u003e Spatiotemporal epidemiology and risk factors of scrub typhus in Hainan Province, China, 2011-2020. \\u003cem\\u003eOne health (Amsterdam, Netherlands)\\u003c/em\\u003e \\u003cstrong\\u003e17\\u003c/strong\\u003e, 100645 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003e孙烨, 方., 曹务春. 中国人民解放军军事医学科学院 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003eTraub, R. \\u0026amp; Wisseman Jr, C.L. The ecology of chigger-borne rickettsiosis (scrub typhus). \\u003cem\\u003eJournal of medical entomology\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, 237-303 (1974).\\u003c/li\\u003e\\n\\u003cli\\u003eSantib\\u0026aacute;\\u0026ntilde;ez, P., Palomar, A.M., Portillo, A., Santib\\u0026aacute;\\u0026ntilde;ez, S. \\u0026amp; Oteo, J.A. The role of chiggers as human pathogens. \\u003cem\\u003eAn overview of tropical diseases\\u003c/em\\u003e \\u003cstrong\\u003e1\\u003c/strong\\u003e, 173-202 (2015).\\u003c/li\\u003e\\n\\u003cli\\u003eChaisiri, K., Cosson, J.-F. \\u0026amp; Morand, S. Infection of rodents by Orientia tsutsugamushi, the agent of scrub typhus, in relation to land use in Thailand. \\u003cem\\u003eTropical Medicine and Infectious Disease\\u003c/em\\u003e \\u003cstrong\\u003e2\\u003c/strong\\u003e, 53 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eXu, G., Walker, D.H., Jupiter, D., Melby, P.C. \\u0026amp; Arcari, C.M. A review of the global epidemiology of scrub typhus. \\u003cem\\u003ePLoS neglected tropical diseases\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, e0006062 (2017).\\u003c/li\\u003e\\n\\u003cli\\u003eBlacksell, S.D.\\u003cem\\u003e, et al.\\u003c/em\\u003e Underrecognized arthropod-borne and zoonotic pathogens in northern and northwestern Thailand: serological evidence and opportunities for awareness. \\u003cem\\u003eVector borne and zoonotic diseases (Larchmont, N.Y.)\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, 285-290 (2015).\\u003c/li\\u003e\\n\\u003cli\\u003eSharma, R. Scrub typhus: prevention and control. \\u003cem\\u003eJK science\\u003c/em\\u003e \\u003cstrong\\u003e12\\u003c/strong\\u003e, 91 (2010).\\u003c/li\\u003e\\n\\u003cli\\u003ePerrone, C.\\u003cem\\u003e, et al.\\u003c/em\\u003e Community engagement around scrub typhus in northern Thailand: a pilot project. \\u003cem\\u003eTransactions of The Royal Society of Tropical Medicine and Hygiene\\u003c/em\\u003e \\u003cstrong\\u003e118\\u003c/strong\\u003e, 666-673 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eRosenberg, R. Drug-resistant scrub typhus: Paradigm and paradox. \\u003cem\\u003eParasitol Today\\u003c/em\\u003e \\u003cstrong\\u003e13\\u003c/strong\\u003e, 131-132 (1997).\\u003c/li\\u003e\\n\\u003cli\\u003eMohapatra, R.K.\\u003cem\\u003e, et al.\\u003c/em\\u003e Linking the increasing epidemiology of scrub typhus transmission in India and South Asia: Are the varying environment and the reservoir animals the factors behind? \\u003cem\\u003eFrontiers in Tropical Diseases\\u003c/em\\u003e\\u003cstrong\\u003e5\\u003c/strong\\u003e, 1371905 (2024).\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":true,\"hideJournal\":false,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-portfolio\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Nature Portfolio\",\"twitterHandle\":\"\",\"acdcEnabled\":false,\"dfaEnabled\":false,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false},\"keywords\":\"\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-6251403/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-6251403/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eScrub typhus, an acute febrile illness caused by \\u003cem\\u003eOrientia tsutsugamushi\\u003c/em\\u003e, has emerged as a significant public health concern, expanding beyond its traditional endemic region, the \\\"tsutsugamushi triangle\\\" in the Asia-Pacific. Despite its increasingly global distribution, comprehensive spatial assessments of scrub typhus risk remain sparse.\\u003c/p\\u003e \\u003cp\\u003eAn exhaustive assembly of 56,093 unique human scrub typhus occurrence records worldwide was undertaken from published literature and national surveillance datasets. Covering 27 countries/regions, these records were combined with 28 climatic, geographic, and socio-economic covariates environmental covariates using an ensemble machine learning modelling approach, capturing possible nonlinear effects and complex interactions, to map the probability of occurrence at 5\\u0026times;5 km resolution globally. This approach involved stacking of three sub-models (generalized additive models, boosted regression trees and random forest). Environmental suitability for scrub typhus was found to be highest in moderate to tropical climates, notably extending beyond the classic \\\"tsutsugamushi triangle\\\" into large sections of Central and South America, Central and West Africa. Approximately 2.5\\u0026nbsp;billion people (95% CI: 2.43\\u0026ndash;2.69\\u0026nbsp;billion) are estimated to be currently living in environmentally suitable areas within countries or regions where human cases of scrub typhus have already been confirmed. This number increases to 4.4\\u0026nbsp;billion people (95% CI: 3.86\\u0026ndash;4.90\\u0026nbsp;billion) if countries without confirmed cases are included.\\u003c/p\\u003e \\u003cp\\u003eThis data assembly and modelled environmental suitability surface provide novel insights into the potential public health impact of scrub typhus. This may serve as a catalyst for broader discussions regarding the neglected global impact of this disease, the need to improve public awareness, drug, and vector control methods, and lead to further burden assessment. The study highlights key data gaps, particularly in regions with limited surveillance and accessibility of healthcare facilities, and emphasizes the need for future research in the context of ongoing climate and environmental changes, which may further alter the global distribution of scrub typhus.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Mapping the global distribution of and environmental suitability for scrub typhus\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-03-25 19:01:00\",\"doi\":\"10.21203/rs.3.rs-6251403/v1\",\"editorialEvents\":[],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-communications\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"NCOMMS\",\"sideBox\":\"Learn more about [Nature Communications](http://www.nature.com/ncomms/)\",\"snPcode\":\"\",\"submissionUrl\":\"https://mts-ncomms.nature.com/\",\"title\":\"Nature Communications\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"Nature Communications\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false}}],\"origin\":\"\",\"ownerIdentity\":\"9828a0f3-e153-4fcf-aa00-93527b372728\",\"owner\":[],\"postedDate\":\"March 25th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"under-review\",\"subjectAreas\":[{\"id\":45963328,\"name\":\"Health sciences/Diseases/Infectious diseases/Bacterial infection\"},{\"id\":45963329,\"name\":\"Health sciences/Risk factors\"},{\"id\":45963330,\"name\":\"Health sciences/Medical research/Epidemiology\"}],\"tags\":[],\"updatedAt\":\"2026-03-13T22:16:00+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2025-03-25 19:01:00\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-6251403\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-6251403\",\"identity\":\"rs-6251403\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}