Emerging Infectious Diseases Review for Lecture PDF

REVIEWS Infectious disease in an era of global change Rachel E. Baker 1,2 ✉, Ayesha S. Mahmud3, Ian F. Miller 1,4, Malavika Rajeev1, Fidisoa Rasambainarivo1,2,5, Benjamin L. Rice1,6, Saki...

REVIEWS Infectious disease in an era of global change Rachel E. Baker 1,2 ✉, Ayesha S. Mahmud3, Ian F. Miller 1,4, Malavika Rajeev1, Fidisoa Rasambainarivo1,2,5, Benjamin L. Rice1,6, Saki Takahashi7, Andrew J. Tatem8, Caroline E. Wagner9, Lin-​Fa Wang 10,11, Amy Wesolowski12 and C. Jessica E. Metcalf1,13 ✉ Abstract | The twenty-​first century has witnessed a wave of severe infectious disease outbreaks, not least the COVID-19 pandemic, which has had a devastating impact on lives and livelihoods around the globe. The 2003 severe acute respiratory syndrome coronavirus outbreak, the 2009 swine flu pandemic, the 2012 Middle East respiratory syndrome coronavirus outbreak, the 2013–2016 Ebola virus disease epidemic in West Africa and the 2015 Zika virus disease epidemic all resulted in substantial morbidity and mortality while spreading across borders to infect people in multiple countries. At the same time, the past few decades have ushered in an unprecedented era of technological, demographic and climatic change: airline flights have doubled since 2000, since 2007 more people live in urban areas than rural areas, population numbers continue to climb and climate change presents an escalating threat to society. In this Review, we consider the extent to which these recent global changes have increased the risk of infectious disease outbreaks, even as improved sanitation and access to health care have resulted in considerable progress worldwide. In premodern times, colonization, slavery and war led spread quickly, aided by global connectivity and shifted to the global spread of infectious diseases, with dev- ranges owing to climate change (Fig. 1d). astating consequences (Fig. 1a). Human diseases such Here, we review how recent anthropogenic climatic, as tuberculosis, polio, smallpox and diphtheria circu- demographic and technological changes have altered lated widely, and before the advent of vaccines, these the landscape of infectious disease risk in the past two diseases caused substantial morbidity and mortality. decades. In terms of climate change, we consider both At the same time, animal diseases such as rinderpest the influence of recent warming and projected future spread along trade routes and with travelling armies, changes. For demographic change, we include trends with devastating impacts on livestock and dependent such as urbanization (Fig. 1b) , population growth, human populations1. However, in the past two decades, land-​use change, migration, ageing and changing birth medical advances, access to health care and improved rates. For technological changes, we primarily consider sanitation have reduced the overall mortality and mor- advances that enable cheaper, faster global travel and bidity linked to infectious diseases, particularly for trade (Fig. 1b), as well as improved health care. We do not lower respiratory tract infections and diarrhoeal dis- explicitly address economic change; however, economic ease (Fig. 1d). The swift development of the severe acute changes, including economic development, are crucial respiratory syndrome coronavirus 2 (SARS-​C oV-2) drivers of these three factors: climate, demography and vaccine speaks to the efficacy of modern science in technology. We also do not explicitly discuss natural rapidly countering threats from emerging pathogens. drivers of pathogen evolution or biological processes Nevertheless, infectious disease burden remains sub- unless they interact with human-​driven global change. stantial in countries with low and lower-​middle incomes, New infections chart a pathway beginning with emer- while mortality and morbidity associated with neglected gence, followed by local-​scale transmission, movement tropical diseases, HIV infection, tuberculosis and beyond borders and possible global-​scale spread. Global malaria remain high. Moreover, deaths from emerging changes may differentially affect the risk of emergence, and re-​emerging infections, in comparison with seasonal the dynamics of disease within a local population and the ✉e-​mail: racheleb@ and endemic infections, have persisted throughout the global spread of diseases between populations. We pro- princeton.edu; cmetcalf@ princeton.edu twenty-​first century (Fig. 1c). This points to a possible vide an overview of each step, first considering features https://doi.org/10.1038/ new era of infectious disease, defined by outbreaks of of recent global change that have altered the risks of s41579-021-00639-​z emerging, re-​emerging and endemic pathogens that spillover of viral, fungal, bacterial and apicomplexan NATure RevieWS | MICrOBIOlOGy volume 20 | April 2022 | 193 0123456789();: Reviews (malaria) infections into human populations, then detail- Fig. 1 | Human connectivity and infectious disease out- ◀ ing how spread within human populations, driven by the breaks in premodern and modern times. a | Examples of seasonal dynamics of transmission, may be impacted by epidemic periods associated with different eras of human global change, of relevance to both emergent and estab- transportation (land, maritime and air travel) are shown. Overland trade networks and war campaigns are thought lished pathogens. Finally, we consider changes to the to have contributed to multiple epidemics in the Mediter­ drivers of global spread, focusing in particular on travel, ranean in late classical antiquity (green), beginning with migration and animal and plant trade. the Antonine plague, which reportedly claimed the life of the Roman emperor Lucius Verus125–128. Maritime trans- Pathogen emergence into human populations portation (red and grey) leading to European contact with Recent decades have seen repeated pathogen emergence the Americas and the subsequent Atlantic slave trade from wild or domestic animal reservoirs into human resulted in the importation of Plasmodium falciparum populations, from HIV-1 and HIV-2, to the 1918 influ- malaria and novel viral pathogens129. In modern times, air enza virus, to Middle East respiratory syndrome corona­ travel (purple) resulted in the importation of severe acute virus, to SARS-​CoV-2 (refs2–4). For a novel pathogen respiratory syndrome (SARS) coronavirus to 27 countries before transmission was halted130. b | In recent years, to become a threat to human populations, first, con- increases in air travel, trade and urbanization at global (left) tact between humans and the animal reservoir must and regional (right) scales have accelerated, indicating occur; the pathogen must either have or evolve (Box 1) ever more frequent transport of people and goods the capacity for human-​to-​human transmission5; and between growing urban areas (source World Bank). c | Log finally, this human-​to-​human transmission must enable deaths from major epidemics in the twenty-​first century expansion of the pathogen’s geographical range beyond (source World Health Organization). d | Disability-​adjusted the zone of spillover. Recent global changes have affected life years lost from infectious diseases (source Our World each of these steps. in Data). MERS, Middle East respiratory syndrome; Patterns of contact between human and wildlife reser­ NTD, neglected tropical disease. voirs have increased as human populations move into previously unoccupied regions. Population growth and agriculture and its intensification may create conditions agricultural expansion, coupled with increasing wealth that favour pathogen circulation within domestic animal and larger property sizes, are driving factors for these (or plant) reservoirs via high-​density farming practices9. interactions and the resulting habitat destruction. This Beyond creating opportunities for emergence of prob- may occur alongside behaviours that increase the poten- lematic livestock pathogens, this could also increase tial for spillover, such as consumption of wild meat6, or opportunities for evolution of novel variants of risk intensifying contact between wild and domestic animal to humans in domestic animal reservoirs. This may hosts. For example, Nipah virus has been identified in occur alongside increasing risk to workers interacting several bat populations, particularly flying foxes, but with animal populations10 as a result of work practices. in 1999 caused a severe disease outbreak in Malaysia, Global increase in the demand for and resulting intensi- primarily among pig farmers7. It is hypothesized that the fication of meat production will importantly drive these spillover of Nipah virus from bats to pigs was driven by processes, and associated use of antibiotics in domestic three factors related to global change: pig farms expand- animals has the potential to select for resistant strains of ing into the bat habitat; intensification of pig farming, bacteria with potential to affect human health11. leading to a high density of hosts; and international The nature of human populations that are exposed trade, leading to the spread of the infection among other to potential spillover is also changing. For example, the pig populations in Malaysia and Singapore8. Expanding elimination of smallpox led to the cessation of smallpox vaccination, which may have enabled the expansion of Author addresses monkeypox12. More generally, globally ageing popula- tions may provide an immune landscape that is more at 1 Department of Ecology and Evolutionary Biology, Princeton University, Princeton, risk of spillover, as ageing immune landscapes are less NJ, USA. capable of containing infectious agents13. The intersec- 2 Princeton High Meadows Environmental Institute, Princeton University, Princeton, tion between declining function of immunity at later NJ, USA. 3 Department of Demography, University of California, Berkeley, Berkeley, CA, USA. ages14 and globally ageing populations may increase the 4 Rocky Mountain Biological Laboratory, Crested Butte, CO, USA. probability of pathogen emergence, but this remains 5 Mahaliana Labs SARL, Antananarivo, Madagascar. conjectural and an important area for research. The 6 Madagascar Health and Environmental Research (MAHERY), Maroantsetra, Madagascar. changing global context may allow existing human path- 7 EPPIcenter Program, Division of HIV, ID, and Global Medicine, Department of Medicine, ogens to both evolve novel characteristics and expand University of California, San Francisco, San Francisco, CA, USA. in scope. Selection for drug resistance now occurs 8 WorldPop, School of Geography and Environmental Science, University of Southampton, worldwide, and antibiotic resistance has and will evolve Southampton, UK. repeatedly15. As with antibiotic resistance, rapid global 9 Department of Bioengineering, McGill University, Montreal, Quebec, Canada. spread is commonplace for antimalarial resistance 10 Programme in Emerging Infectious Diseases, Duke-​NUS Medical School, Singapore, following evolution16. Singapore. 11 Duke Global Health Institute, Duke University, Durham, NC, USA. Climate change may play a role in the risk from path- 12 Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Johns ogen spillover. Changing environmental conditions can Hopkins University, Baltimore, MD, USA. alter species range and density, leading to novel interac- 13 Princeton School of Public and International Affairs, Princeton University, Princeton, tions between species, and increase the risk of zoonotic NJ, USA. emergence17. A series of compounded environmental 194 | April 2022 | volume 20 www.nature.com/nrmicro 0123456789();:  Reviews a Post-Columbus contact and Classical antiquity trade and war: Transatlantic slave trade International air travel: European colonization: Antonine plague (second century) and European colonization: SARS epidemic smallpox, measles and other Plague of Cyprian (third century) P. falciparum malaria (2002–2004) diseases (fifteenth to Justinian plague (sixth century) (sixteenth to nineteenth eighteenth century) century) b c 4 Ebola virus disease Exports of goods and services Air travel passengers COVID-19 1.0 Emerging Ebola virus disease 3 Re-emerging Swine flu 8 2 Zika virus disease 0.5 Measles 6 Cholera (billion) Deaths from major 1 Dengue MERS SARS epidemics (log) 0.0 4 1970 1980 1990 2000 2010 2020 1970 1980 1990 2000 2010 2020 2 25 10.0 20 0 (trillion current US$) 7.5 2000 2005 2010 2015 2020 15 5.0 Year 10 2.5 5 d 0 0.0 Disease category 1970 1980 1990 2000 2010 2020 1970 1980 1990 2000 2010 2020 Disability-adjusted life years 3 Other Urban population (billion) 4 Tuberculosis 2 (billion) 1.0 HIV/AIDS 3 NTDs and malaria 1 0.5 Diarrhoea and lower 2 respiratory tract infections 0 0.0 1990 2000 2010 2020 1970 1980 1990 2000 2010 2020 1970 1980 1990 2000 2010 2020 World Sub-Saharan Africa East Asia and Pacific Latin America and Caribbean Europe and Central Asia North America Middle East and North Africa factors, including a long period of drought followed by evidence suggests that populations of the black flying fox extreme precipitation, is hypothesized to have driven an in Australia, a key reservoir of Hendra virus, have moved upsurge in rodent populations causing the emergence 100 km southward in the past 100 years owing to cli- of pulmonary hantavirus in 1993 (ref.18). Similarly, matic changes. This shifting range likely caused Hendra NATure RevieWS | MICrOBIOlOGy volume 20 | April 2022 | 195 0123456789();: Reviews virus to spill over into southern horse populations, and connected urban areas are potential hot spots for the these horses subsequently infected humans19,20. Patterns rapid spread of diseases such as COVID-19 and SARS, of change are likely occurring in other bat populations and cities can serve as a catalyst for rapid local and global globally but remain understudied — a clear cause for transmission. concern given the crucial role bat populations play as a reservoir host for several high-​fatality pathogens21. Local-​scale disease dynamics Rapid rates of urbanization in low-​income and Emerging, re-​emerging and endemic pathogens in middle-​income countries, and the increase in popula- human populations may exhibit distinct dynamic pat- tions residing in crowded, low-​quality dwellings, have terns of spread at the local scale. These patterns will created new opportunities for the emergence of infec- be governed by demographic factors, including the tious diseases (Fig. 2). Urbanization has promoted the effects of human behaviour on transmission (for exam- emergence and spread of arboviral diseases such as ple, school terms drive transmission of many child- dengue, Zika virus disease and chikungunya, which hood infections26 and sex-​specific travel patterns may are transmitted by Aedes aegypti and Aedes albopictus result in higher burdens of chikungunya in women in mosquitoes that are well adapted to urban areas22–24. Bangladesh27) and immunity (which, for immunizing Population density appears correlated with the pref- infections such as measles and rotavirus infection, is, in erence of Ae. aegypti for human odour, and hence the turn, shaped by replenishment of susceptible individuals evolution of human-​biting — the transmission pathway via births28,29 and depletion by vaccination where vac- for arboviral disease24. However the role of urbanization cines are available30). Transmission may also be affected in vector-​borne disease spread is complex: the prefer- by climatic variables acting spatially or over the course ence of the Anopheles spp. vector for rural environments of the year in line with seasonal fluctuations31,32. Recent may have led to a decline in the prevalence of malaria global changes have affected each of these drivers of in urbanizing regions25. Nevertheless, dense and highly local-​scale dynamics (Fig. 3). As school attendance not only modulates transmis- sion of childhood infections26 but also shapes human Box 1 | Global change and evolution of hosts and pathogens mobility33, dramatic increases in rates of school attend- Mutations constantly arise in the genomes of all species, from viruses to elephants. ance globally thus have the potential to substantially alter Some genetic changes may have no observable effects on fitness (and thus will be the dynamics of many infections. That this has yet to selectively neutral), but can be used to track pathogen spread; for example, to trace be documented is perhaps in part because this change the impacts of global connectivity on an outbreak70. Some genetic changes will affect has happened alongside expansion of access to vaccines disease phenotypes, potentially increasing the transmissibility, virulence or immune that protect children against many of the relevant infec- escape of a pathogen lineage133. The degree to which such mutations increase in tions, as well as global declines in birth rates, which also frequency or spread geographically will depend on the degree to which they increase fitness, as well as pathogen population dynamics, which may be modulated by the facilitate control efforts by diminishing the size of the global change context. Increases in the density and geographical distribution of susceptible pool34. If the burden of disease is age specific, susceptible hosts (whether they be people, crops or livestock) may provide greater the intersection between immunity and shifting demo­ opportunity for novel variants to emerge9 simply by amplifying pathogen populations graphy may be particularly marked: declining birth rates and thus circulating mutations. While understanding the nuance of cross-​scale translate into a smaller pool of susceptible individuals selection (that is, how the selective context of the individual host translates into the and thus infected individuals, reducing the overall rate selective context at the scale of populations) remains a challenging frontier134, it is at which susceptible individuals become infected, and likely that ageing populations or the presence of immunosuppressive pathogens thus increasing the average age of infection or disease, as might further modulate selection pressures. Indeed, it has been suggested that the reported for dengue in Thailand35 and rubella in Costa emergence of more transmissible or less immune-​vulnerable variants of severe acute Rica36 as these countries went through the demographic respiratory syndrome coronavirus 2 (SARS-​CoV-2) was enabled in part by selection processes occurring during chronic infections in immunosuppressed individuals135. transition. Conversely, ageing populations may increase Greater global connectivity leads to more frequent exchange of this genetic material transmission; for example, longer shedding has been between populations of the same or different species, potentially leading to the erosion suggested with increasing age for SARS-​CoV-2 (ref.37). of evolved or engineered host resistance and increased rates of pathogen evolution136. Demographic changes to population size and density Associated spillover followed by spillback can create scenarios that facilitate via urbanization may also affect dynamics. Influenza, for amplification and potentially selection of problematic pathogen variants137, an issue example, tends to exhibit more persistent outbreaks in highlighted by recent documentation of human to mink to human transmission of more populous, denser urban regions38 (Fig. 2). A similar SARS-​CoV-2 (ref.138). Likewise, increased rates of pathogen importation provide pattern was reported in the early COVID-19 pandemic39. increased opportunities for pathogen populations to evolve the ability to utilize novel If demographic change has importantly altered the con- vectors (as has been observed in the Americas for malaria129). Increased population text of infectious diseases in recent years, arguably an connectivity can also enable pathogens and their vectors to shift to novel host species, from infected mosquitoes travelling on boats or in planes to agricultural pathogens even larger effect is caused by changes in the occurrence being inadvertently relocated. Hosts that have not previously been exposed to such of immunomodulatory infections, which, in turn, may pathogens, and thus have no co-​evolved defences, yet are phylogenetically and/or affect other infections. For example, the emergence of genetically similar to the original host are often most at risk139,140, a fact that makes HIV has amplified the burden of tuberculosis40. Mass homogenization of crops141 or livestock a concern. Novel pathogen introductions can drug administration efforts have reduced helminth have large-​scale population and ecosystem impacts, of which one famous example prevalence, which will have knock-​on effects on the is the extirpation of the American chestnut tree by chestnut blight142. Changes in burden of other infections, such as malaria, which may selection pressure resulting from changes in health-​care strategies (for example, be increased in individuals experiencing a heavy worm introduction of vaccination) may have the potential to select for different pathogen burden41; both will also intersect with the efficacy of characteristics, and could potentially drive the evolution of virulence in pathogens143,144. vaccination programmes42. 196 | April 2022 | volume 20 www.nature.com/nrmicro 0123456789();:  Reviews More persistent outbreaks of respiratory disease in urban settings Air pollution increases disease Global transit networks susceptibility increase risk of importing and exporting pathogens Urban expansion and Improved Poor sanitation and Aedes aegypti and Aedes Anopheles spp. less land-use change lead health-care high population density albopictus adapted to adapted to urban to novel human–wildlife access increase spread of urban settings, breed in areas, possible interactions enteric infections standing water declines in malaria Fig. 2 | Impacts of urbanization on infectious disease. Interactions between urbanization and infectious disease are complex, with increased urbanization driving both positive and negative changes to global disease burden. The climate plays a key role in driving the local-​scale thermotolerance may have enabled the pathogen to seasonal dynamics of many infectious diseases, which jump from its environmental habitat into an interme- may thus be altered by global change in climatic diary avian host, given the higher body temperatures of conditions43,44. Considering these impacts requires avian fauna, before infecting humans52. recog­nizing that interactions with climate differ by path- Demographic change and technological changes may ogen type. For directly transmitted infections, the role alter a host’s interaction with the environmental reser- of climate is revealed by marked latitudinal gradients voir. Cholera, caused by the bacterial pathogen Vibrio in epidemic timing32,45. Several respiratory pathogens, cholerae, persists in the environment, particularly in including influenza virus, are more highly seasonal in aquatic settings. Changes to environmental conditions, temperate climates and exhibit greater year-​round including elevated sea temperatures, lead to increased persistence in tropical locations32,46. Climate change reproduction of the pathogen and local epidemics53, with is expected to lead to an expansion of these tropical clear links to longer-​term climate phenomena such as patterns, with possible implications for pathogen El Niño54. However, improved sanitation lowers the risk evolution43,47. At the individual level, susceptibility to of exposure to V. cholerae and has led to a decline of the respiratory viral infections may be impacted by expo- disease in many locations53. sure to local air pollution, which is a concern for rapidly For vector-​transmitted diseases, biological traits of urbanizing locations, where urban air pollution may both the vector and the pathogen may be sensitive to disproportionately affect low-​income communities and climate. Many transmission-​related life cycle traits of communities of colour48,49. For example, non-​Hispanic the mosquito (biting rate, adult lifespan, population size Black and Hispanic populations in the USA were and distribution) and the pathogen (extrinsic incuba- found to have higher exposure to certain PM2.5 com- tion rate) are temperature sensitive, and oviposition ponents than non-​Hispanic white populations49. At the patterns depend on water availability55. Consequently, same time, globally, a move to an urban location may the geographical range for dengue, malaria and other bring benefits in terms of increased access to health vector-​borne diseases56–58 is affected by the local cli- care (Fig. 2). mate, and there is substantial effort to understand how For some bacterial and fungal diseases, climatic these ranges may change with climate change59–61. For changes may affect the pathogen’s environmental reser­ certain vector-​borne diseases such as Zika virus dis- voir. Incidence of coccidioidomycosis (valley fever), ease, climate change may lead to an expanded range62. caused by inhalation of fungal spores of Coccidioides However, for other diseases, such as malaria, climate spp., is expected to increase with climate change as the change may shift the spatial range of the infection to region with optimal conditions for fungal spore produc- higher latitudes63. As ever, the footprint of human inter- tion expands50. Climate change may also have played ventions may loom larger than these changes in local a role in the emergence of the drug-​resistant fungal conditions25. pathogen Candida auris. C. auris emerged in several At the local scale, one of the strongest footprints El Niño continents at the same time and has been shown to detectable on the dynamics of many endemic infections A correlated series of climate events associated with the have increased thermotolerance compared with other in recent years is declines in incidence associated with warm phase of the El Niño closely related fungal species, which perhaps evolved access to vaccinations64. However, the introduction Southern Oscillation cycle. in response to global warming 51,52. This increased of a vaccine does not imply immediate elimination. NATure RevieWS | MICrOBIOlOGy volume 20 | April 2022 | 197 0123456789();: Reviews As vaccination coverage increases, measles outbreaks, for technological change have increased the trade of plants instance, follow a pathway towards elimination defined and animals. At the same time, an increasingly urban by declines in mean incidence but high variability in population is better connected than ever before to global outbreak size34. Imperfect vaccine coverage may allow travel networks (Fig. 4). These changes to global con- population susceptibility to increase such that substan- nectivity will present unique risk factors for infectious tial outbreaks can occur if the disease is reintroduced; disease spread, enabling pathogens to travel further and for example, the 2018 measles outbreak in Madagascar, faster than ever before. which led to more than 100,000 cases65. Improved surveillance of the landscape of population immunity, International travel. The late twentieth century and the via serological surveys, could help determine gaps in early twenty-​first century have been marked by techno- vaccination coverage66. logical developments enabling ever swifter movement of people and pathogens over large distances — from Global spread trains to planes, and an expanding international airline As local conditions alter demographically, or as a result network (Fig. 4). The total number of airline passengers of climate change potentially expanding the range of doubled from just below two billion in 2000 to more locations suitable to a particular pathogen or vector, than four billion in 2019 (Fig. 1b). This rampant increase increased global connectivity will enable pathogens to in global connectivity brings with it new risks from reach these new environments more rapidly (Figs 3,4). emerging pathogens (Box 2). However, many endemic Here, we review the impact of global change on three pathogens also circulate via transit routes: seasonal influ- forms of global connectivity — international travel, enza circulation in the USA can be predicted by flight human migration and local-​s cale mobility, and the patterns67,68, with evidence that flight bans following the international trade of animals, animal products and events of 9/11 caused a delayed outbreak, and a pro- plants — while considering the impact on infectious longed influenza season within the USA as measured disease risk. Technological change over the past two dec- by a 60% increase in the time to transnational spread68. ades has dramatically lowered the cost of international Similarly, rapid global air travel is expected to have travel, while demographic change has led to heightened played a key role in the global spread of SARS-​CoV-2. demand for inexpensive flights (Fig. 1b). Demographic Genetic analyses demonstrate multiple introductions of and climatic drivers have altered patterns of local mobil- SARS-​CoV-2, driven by air travel, in the Middle East69, ity and regional migration, while rising demand and northern California70 and Brazil71. Disease dynamics β×S×I γ×I Pathogen S I R emergence Global spread Pathogen evolution Climatic change Drives range shifts for Affects transmission and Affects the geographical range reservoir species susceptibility of vectors Technological change Transportation Improved global Air transit and high-speed rail surveillance affect pace and range of spread Health care Vaccination affects Improved care reduces burden dynamics Demographic change Population growth and Increased contact with Population numbers affect Larger population travelling land use reservoir species evolution, birth rates affect dynamics Urbanization Depends on species Density affects contact rate Urban population more connected Ageing Imunosenescence affects Ageing population increases Possible larger burden spillover risk transmission Fig. 3 | Effects of climatic, technological and demographic change on disease emergence, dynamics and spread. The table summarizes select recent global changes (rows) and their impacts on disease emergence, local-​scale dynamics and global spread (columns). An example susceptible (S), infected (I), recovered (R) model is shown, where β represents the transmission rate and γ is the recovery rate. 198 | April 2022 | volume 20 www.nature.com/nrmicro 0123456789();:  Reviews a Fig. 4 | Mapping changes to travel and climate. a | The global international air travel network expanded substantially from 1933 to 2020 (data from WorldPop and ref.131). b | Average monthly maximum temperature in 1970–2000) and difference between 2070–2100 and 1970–2000 averages (data from WorldClim, Shared (Recent) increasing connectivity Socioeconomic Pathway 3 (SSP3)). c | Population projections under SSP3 in 2010 and relative population change projected until 2100 (source NASA Socioeconomic Data 1933 and Applications Center (ref.132)). Part a adapted with permission from ref.131, OUP. International travel can lead to the global spread of vector-​borne diseases via the introduction of new vec- tors into regions with suitable environmental conditions or the introduction of new pathogens into native and invasive vector populations. Historically, vectors have been introduced via trade routes: ships are thought to have been key to the global dispersal of Ae. aegypti and 2020 Ae. albopictus, which then became established in loca- tions with appropriate environmental conditions72,73. b Anopheles gambiae, the primary vector of malaria in Africa, was introduced into Brazil in the 1930s and became established in a region with a climate similar to ºC 40 that of its native Kenya74. Although malaria was already 30 endemic in Brazil at the time, An. gambiae proved a 20 much more effective vector, leading to a severe outbreak (Future) changing climate 10 and a costly (but successful) eradication campaign73. 0 –10 There has been relatively little documented evidence –20 of the introduction of new vectors via air travel. This is –30 2000 likely due to the low probability of vectors surviving the flight, and disembarking in a suitable region, in sufficient numbers to establish and drive an epidemic75. However, cases of ‘airport malaria’, that is, malaria transmitted ∆ºC 10 within international airports, even outside endemic 8 regions, are rare but becoming more common76. A more feasible scenario is that air travel can bring an 6 infected human host into contact with a native or inva- 4 sive vector population that then establishes local trans- 2 mission. Climate change has driven a shift in the range of 0 2100 − 2010 several key vectors, which may make this introduction more likely. The range of the biting midge Culicoides imicola, Log (population) a vector of bluetongue virus, which causes disease in c 7 ruminants, has expanded over the past few decades from 6 sub-​Saharan Africa and the Middle East into Europe, 5 bringing a wave of bluetongue epidemics77. Following 4 this expansion, bluetongue virus then spread outside the range of C. imicola into native populations of Culicoides 3 spp. in more northerly regions of Europe. In terms of (Future) changing population 2 air travel, the 2015 Zika virus disease epidemic in the 1 Americas may provide a recent example of a pathogen 2010 spreading into a susceptible vector population, likely 0 facilitated by high connectivity78. Zika virus is thought ∆ (population) to have been introduced to Brazil from French Polynesia 12 and vectored by Aedes spp., although the volume of air 10 travel during this period makes it almost impossible to 8 conclusively determine the origin78. Similarly, it is hard 6 to pinpoint the pathway via which West Nile virus was 4 introduced into the USA in the 1990s; however, trans- 2 port by either shipping (transporting vectors) or aircraft 0 (transporting a human host) is likely79. After introduc- tion, West Nile virus spread in the native Culex spp. 2100 − 2010 mosquito population. More broadly, climate change NATure RevieWS | MICrOBIOlOGy volume 20 | April 2022 | 199 0123456789();: Reviews Box 2 | Will there be another pandemic like COVID-19? COVID-19 has had an unprec- 12 edented impact on both 36 h human lives and our society, and we will likely be dealing Presymptomatic spread 24 h Incubation period (time to symptoms (days)) with the con­sequences for 10 decades to come. As we 12 h reckon with these con­ sequences, one concern is that a suite of global changes 8 has increased the risk from emerging pathogens, such that pandemics similar to COVID-19 could be a more 6 SARS-CoV-2 Ebola virus frequent occurrence. However, there are biological features of severe acute respiratory SARS-CoV 4 syndrome coronavirus 2 (SARS-​CoV-2) that have made the pathogen distinctly diffi- Postsymptomatic spread cult to control, primarily the 2 virus’s ability to spread asym­ Influenza virus ptomatically and presympto- matically. Many pathogens 2 4 6 8 10 12 do not exhibit these features, Latent period (time to infectiousness (days)) which may be a cause for cautious optimism going forward. The expansion of regional and global air travel, along with the increasing development of high-​speed railway networks, has resulted in a substantial degree of connectivity between human populations73. At the same time, land-​use change and climate change may have increased the risk of pathogen emergence. In combination, these drivers imply an era where pathogens are more likely to emerge, and more likely to spread globally on emergence. However, while the last century bore witness to several pandemics (Fig. 1), SARS-​CoV-2 is unrivalled in its rapid, global reach. A key question is why SARS-​CoV-2 was so successful at spreading globally and whether this was due to recent increases in global connectivity as opposed to epidemiological and biological characteristics of the virus itself145. A clear distinction between SARS-​CoV-2 and other recently emerged pathogens (for example, SARS-​CoV and Ebola virus) is that an individual infected with SARS-​CoV-2 may become infectious before developing symptoms146. This presents a unique challenge from a disease control perspective. A standard approach for limiting the onward spread of a new outbreak is to isolate infected individuals when they show symptoms. Case isolation proved successful in mitigating earlier SARS147 and Ebola virus disease148 outbreaks. However, symptoms for SARS-​CoV-2 infection likely occur after an individual is already infectious146. This possible presymptomatic spread limits the efficacy of case isolation interventions as by the time the infected individual is isolated, the person may have already spread the pathogen to others149. In the figure, we plot the time to infectiousness (latent period) against the time to symptom onset (incubation period) for four pathogens that have caused severe outbreaks in recent decades. When the latent period equals the incubation period (dashed line in the figure), symptoms occur at a similar time to infectiousness (for example, influenza). The shaded region to the right of this line in the figure indicates possible presymptomatic spread, which may be uniquely difficult to control. The 2–3-day delay between infectiousness and symptom onset provides ample time for long-​distance spread of the disease, given current transport networks (see the figure). Control policies, such as testing before travel, provide a more effective option in this context, yet developing and distributing a test takes time, during which time the disease may spread rapidly. The good news is that this presymptomatic spread appears somewhat unique to SARS-​CoV-2, at least compared with other acute infections such as influenza, SARS and Ebola virus disease (Fig. 4). In comparison, asymptomatic spread explains some of the difficulty in controlling acquired immunodeficiency syndrome before antiretroviral measures were available. complicates the picture in terms of possible future intro- and one that has undergone considerable upheaval in ductions. As the range of locations with environmental the modern era. It is estimated that globally the number suitability for certain vector species changes, successful of international migrants, those who intentionally relo- introductions of pathogens into these vector populations cate to a country other than their birth country, is almost may become more likely80. At the same time, changes 272 million, representing 3.5% of the world’s population. to population structure (for example, via urbanization) By contrast, temporary migration, often considered ‘sea- may alter the suitability of an environment for vector sonal migration’, is driven largely by economic patterns, reproduction (Fig. 2). including agricultural seasons that require short peri- ods of intense labour. The rate of migration continues Migration and local mobility. Human migration is an to increase owing to both social, economic, political and intrinsic component of population dynamics driven by environmental drivers in origin countries and economic socio-​economic, political and environmental factors, opportunities, physical safety and security in destination 200 | April 2022 | volume 20 www.nature.com/nrmicro 0123456789();:  Reviews countries. Projections for migration are unclear, with the Combined with the intensification of trade at the global UN projecting stable rates after 2050 (ref.81). However, scale, this pattern has driven a rise in long-​distance climate change will likely provide an escalating push transmission and disease emergence98,99. Trade drives factor, with sea level rise and extreme weather events the emergence of novel plant diseases by creating novel leading to forced migration from exposed regions82. interactions between hosts and pathogens100. One path- Given the movement of people between countries, way through which this can occur is the introduction of there remain risks of introduction of infectious diseases, novel pathogens to native plants. For example, Xylella including those common and uncommon in the country fastidiosa, a generalist bacterium vectored by xylem-​ of migration83. It is possible for a infectious disease com- feeding insects, was introduced into Europe in 2013 from mon in the source country, such as latent tuber­culosis, the USA, likely as a result of trade. In Italy, X. fastidiosa is malaria, viral hepatitis and infection with intestinal causing an ongoing epidemic of ‘olive quick decline syn- parasites, to be imported via this mechanism84–86. For drome’, resulting in severe losses of an economically and example, in many destination countries, a large pro- culturally important crop101,102. Trade can also drive the portion of cases of tuberculosis are observed in the emergence of plant disease by introducing novel hosts foreign-​born population. However, the ultimate impact to native pathogens. Eucalyptus rust, a disease caused of these introduction events will depend largely on the by the fungal pathogen Austropuccinia psidii, emerged population-​level susceptibility and environmental suita- when the pathogen transferred from its native South bility for sustained transmission in the destination coun- American hosts in the myrtle family (Myrtaceae) to non-​ try. More importantly, migrant groups often have more native Eucalyptus trees (which also belong to the myrtle limited access to health care, treatment and resources, family) being grown on plantations103. The disease now particularly those displaced, who are often provided threatens to ‘spill back’ into naive endemic Eucalyptus with limited options to safely seek care and treatment87. populations in Australia. Minimizing the impact of these possible disease threats depends on providing appropriate health care to these Animal and animal-​product trade. Animal trade has high-​risk groups that takes into account the multifaceted contributed to multiple outbreaks and emergence events social, political and economic components88. globally, which have had major consequences for the Within-​country population mobility can also play a agricultural sector as a whole and pose substantial risk key role in disease spread; however, it is typically dif- for animal and public health. Large numbers of livestock ficult to track these movements. Aggregated mobile are traded annually between countries and may facilitate phone data are a valuable tool for tracing patterns of local the spread of pathogens. Rift Valley fever, for example, mobility and predicting future outbreaks89. In recent is a zoonotic vector-​borne viral disease causing abor- work, mobility data have been shown to be predictive of tion and high neonatal mortality in domestic ruminants. inequities in COVID-19 burden in the USA90. Similarly, The disease is widespread on the African continent and population mobility was found to predict the spread has recently been detected in Saudi Arabia and Yemen. of the 2011 dengue epidemic in Pakistan91, while local Live cattle movement between East Africa and the travel following the Eid holidays was found to predict Arabian peninsula or from the Union of Comoros to the spread of the chikungunya outbreak in 2017 in Madagascar is thought to have contributed to the intro- Bangladesh92. As the trend of urbanization contin- duction of Rift Valley fever virus and caused outbreaks ues, mobility to and from dense urban centres (that is, in these locations in 2000 (Arabian Peninsula) and 2008 mega­cities) will likely play a future role in local spread (Madagascar)104,105. of infections92. Better tracking of within-​country popu- Additionally, the trade of animal-​derived products lation mobility, using novel data streams, may present an such as meat may enable the movement of patho- opportunity for forecasting future outbreaks93. gens over large distances and between continents. For instance, African swine fever is a highly contagious viral Intensification of animal and plant trade disease affecting several members of the family Suidae, International trade has expanded rapidly in the mod- including domestic pigs and wild boars. Infection by ern era and has been matched by a global proliferation African swine fever virus may result in up to 100% of infectious diseases affecting not only humans but morbidity and mortality in affected pig herds and sub- also animals and plants94,95. Trade drives this pattern stantial economic losses for producers. In 2007, the by facilitating the translocation of hosts and pathogens accidental introduction of African swine fever virus to across the geographical and ecological boundaries that Georgia led to the first outbreak of African swine fever constrain their spread. The economic and environmen- in Europe since the early 1990s106. The virus, which used tal threats posed by trade-​driven infectious diseases of to occur primarily in sub-​Saharan Africa, was allegedly plants and animals are increasingly being recognized, introduced to the Caucasian peninsula through meat and calls for more stringent containment measures have products contaminated with viruses closely related intensified in recent years96,97. to the ones found in Madagascar, Mozambique or Zambia107. Despite efforts to contain the virus, the dis- Plant trade. Deliberate transport of plant products has ease has spread to more than 20 countries in Europe and existed since the emergence of trade. Increases in the Asia108,109. Propagules Pathogen units responsible speed of transport during modern times have allowed Similarly, in recent decades there has been an for infection, such as a fungal more live plant tissue, and as a result more viable patho­ expansion in infections of Vibrio parahaemolyticus — spore or viral particle. gen propagules, to be transported over long distances. a bacterial pathogen found in shellfish and the leading NATure RevieWS | MICrOBIOlOGy volume 20 | April 2022 | 201 0123456789();: Reviews Box 3 | Big data for disease amphibian populations globally112. Conversely, infec- tious diseases also hamper trade, resulting in indirect Recent technological advances in collecting, sharing and processing large datasets, economic losses in affected populations. Foot and from satellite images to genomes, represent a new opportunity to answer critical mouth disease virus is a major reason for trade restric- questions in global health. However, challenges remain, including the uneven tions on livestock. While endemic in certain countries geographical distribution of available data as well as biases in representative sampling. in Asia and Africa, foot and mouth disease virus causes We highlight three areas of future growth. outbreaks in naive populations, resulting in large eco- Serological surveys nomic losses113. While trade is a major driver of patho- Serological surveys detect the presence of antibodies in blood — recent advances in gen spread, food animal production has transformed in testing now enable the detection of exposure to multiple pathogens with use of a small recent history into large-​scale intensified systems with sample of blood150. Serological surveys have attracted attention during the COVID-19 pandemic as a means to track population exposure given under-​reporting, although high-​density, genetically homogenous populations, ideal test performance characteristics differ widely between epidemiological contexts as for pathogen emergence and maintenance114. Critically, well as the choice of assay used151. Historically, serological surveys have been financially animal production systems often serve as the interface and logistically expensive to run, but declining costs are leading to increased between wild and human populations, and multiple availability of serological data. viral spillover events have occurred at this nexus. Nipah Genomic surveillance systems virus spilled over from fruit bats to the domestic pig Genomic surveillance systems are able to characterize and track the emergence of population multiple times before subsequently infect- novel variants (for example, during the COVID-19 pandemic). Undoubtedly these ing humans115. Pandemic variants of human influenza data have enabled the rapid development of diagnostics and vaccines and, when A virus are often the result of reassortment between combined with epidemiological information, are able to provide a more detailed picture human and avian viruses, with both domestic poultry of ongoing transmission dynamics. Efforts to develop national and international genomic and wild birds posited to play a role116–118. A non-​viral surveillance networks are varied but with clear success stories152,153 even in low-resources example is the spillover of antimicrobial-​resistant patho- settings154. However, resource limitations, including sequencing platforms, bioinformatic gens from livestock into humans: intensive antibiotic use pipelines and the regular collection of samples for processing, continue to limit the in industrialized and smallholder livestock production global expansion of sequencing. systems to promote growth and prevent infections has Artificial intelligence and machine learning been linked to the emergence of antibiotic resistance in These techniques are frequently proposed as tools for answering key public health humans119. Tackling emergence and disease spread questions, yet specific use cases remain elusive155. Using these tools to predict viral in animal systems will require rethinking both food emergence, for example, may prove difficult due to microbiological complexities and animal production and global trade of animals. the cost of data collection156, yet could prove valuable for targeting sampling efforts157. In terms of uncovering population-​level drivers of disease transmission, statistical approaches, including machine learning, can be used to leverage novel, and high-volume, A new era of infectious disease data streams. However, more classical, mechanistic models may provide a more robust In recent decades, declines in mortality and morbidity, framework for projecting future outcomes for the disease system under demographic, particularly childhood mortality, have been one of the technological and climatic change. Future work should aim to improve the integration of great triumphs of public health. Greater access to care, machine learning approaches within the traditional mechanistic modelling frameworks such as therapeutics (including antibiotics), improved to rapidly and accurately assess prospective challenges. sanitation and the development of vaccines120 have been core drivers of this progress. Even as medical advances in cause of seafood-​related illness globally. The pathogen the twenty-​first century have spurred advances in popu­ is endemic to regions of the US Pacific Northwest but lation health, inequalities in access to these advances has recently spread to other parts of the USA, Europe remain widespread between and within countries121. and South America110,111. The concerning increase in Reducing inequities in access to health care and improv- V. parahaemolyticus infection is expected to have ing surveillance and monitoring for infectious diseases seve­ral drivers connected to global change. Declines in low-​income and middle-​income countries, and in in sea ice have increased ship traffic through the underserved populations within countries, should be Bering Strait, with cargo ships possibly transporting a priority in tackling pathogen emergence and spread. V. parahaemolyticus in ballast water. At the same time, While life expectancy continues to increase, and life increasing sea temperatures may have increased the years lost to infectious diseases decline, the new threat global environmental suitability for V. parahaemolyticus of infectious disease will likely come from emerging and in the marine environment110. Finally, dispersal of the re-​emerging infections. Climate change, rapid urbaniza- pathogen may have occurred via increasing global trade tion and changing land-​use patterns will increase the risk in shellfish, with evidence suggesting possible dispersal of disease emergence in the coming decades. Climate via Manila clams introduced into Spain from Canada111. change, in particular, may alter the range of global path- This combination of possible drivers speaks to the ogens, allowing infections, particularly vector-​borne complexity of understanding infectious disease risk in infections, to expand into new locations. A continued an era of global change, and the necessity of exploring uptick in global travel, trade and mobility will trans- concurrent changes. port pathogens rapidly, following emergence. However, Transboundary spread of diseases through legal there are counterpoints to this trend: the rapid growth of and illegal trade of live animals may also have impor- connectivity observed in the early twenty-​first century tant consequences for biodiversity on a global scale. may stabilize, and structural changes wrought during Reassortment The mixing of genetic material For example, the amphibian trade contributed to the the COVID-19 pandemic may persist122. Increased of different pathogens within expansion of novel strains of the fungal pathogen genus investment in outbreak response, such as the recent for- an infected cell. Batrachochytrium into naive hosts, devastating wild mation of the WHO Hub for Pandemic and Epidemic 202 | April 2022 | volume 20 www.nature.com/nrmicro 0123456789();:  Reviews Multiplex serology Intelligence, could help mitigate the threat from future endemic infections is won, institutional structures The measurement of antibody emerging infections. In addition, efforts to develop uni- built to address these old enemies can be co-​opted and responses to multiple versal vaccines (that is, vaccines that engender immunity adapted for emerging threats. At the same time, new pathogens simultaneously. against all strains of influenza viruses or coronaviruses, technologies, including advances in data collection for example) could provide a monumental leap forward and surveillance, need to be harnessed (Box 3). There is in tackling present and future infections123. much recent innovation around surveillance, from rein- A changing world requires changing science to eval- terpreting information available from classic tools such uate future risks from infectious disease. Future work as PCR124 to leveraging multiplex serology approaches to needs to explicitly address concurrent changes: how identify anomalies that might suggest pathogen emer- shifting patterns of demographic, climatic and techno- gence, and there is increasing interest in integrating logical factors may collectively affect the risk of path- multiple surveillance platforms (from genomic to case ogen emergence, alterations to dynamics and global data) to better understand pathogen spread. Finally, spread. More forward-​looking research, to contend future research needs to align with a global view of disease with possible future outcomes, is required in addition risk. In an increasingly connected world, the risk from to the retroactive analyses that typically dominate the infectious disease is globally shared. The COVID-19 literature. Increasing attention needs to be paid to patho­ pandemic, including the rapid global circulation of gens currently circulating in both wild and domestic evolved strains, highlights the need for a collaborative, animal populations, especially in cases where agriculture worldwide framework for infectious disease research is expanding into native species’ habitats and, conversely, and control. invasive species are moving into populous regions due to climate change. As the battle against certain long-​term Published online 13 October 2021 1. Roeder, P., Mariner, J. & Kock, R. Rinderpest: the 20. Yuen, K. Y. et al. Hendra virus: epidemiology dynamics insights from a statistical analysis and mathematical veterinary perspective on eradication. Philos. Trans. in relation to climate change, diagnostic tests and modeling. PLoS Med. 6, e1000139 (2009). R. Soc. Lond. B Biol. Sci. 368, 20120139 (2013). control measures. One Health 12, 100207 (2021). 36. Metcalf, C. J. E. et al. Structured models of infectious 2. Karesh, W. B. et al. Ecology of zoonoses: natural and Martin et al. (2018) and Yuen et al. (2021) detail