It was evident from the review of abstracts and evaluation of full papers in the final synthesis that potential CSIs in the arthropod vector-borne category dominated, supporting earlier findings [10, 12]. The increasing importance of vector-borne diseases (VBD) at Northern latitudes is generally due to expansion of the geographical range for important vector species and their vertebrate hosts. In particular, many publications focus on tick-borne diseases (TBDs) in Europe (Fig. 4). The TBDs listed in Table 2, i.e. anaplasmosis, babesiosis, borreliosis and TBE, were all included in the full paper reading and the final results indicated that borreliosis and TBE can be classified as climate-sensitive. This supports findings in several European studies regarding the influence of climate change, i.e. distribution and expansion to higher altitudes, on TBDs, particularly TBE and borreliosis [13, 14]. However, TBDs illustrate how new information may change opinions on the influence of climate change over time. Dufour et al. [7] decided to exclude TBDs from their list of potential CSIs, while including insect-borne diseases (by mosquitoes and midges), since the participating experts were unable to decide on how ticks would react to climate change.
The midge-borne disease bluetongue was also classified as climate-sensitive, supported by studies showing increased impact of bluetongue as higher temperature opens up new geographical areas for both the vectors and the virus [15,16,17]. Lastly, fasciolosis, a parasitic infection affecting both wildlife and domesticated animals [18], was classified as climate-sensitive.
The present study included a high proportion (74%) of zoonotic infections. It has been suggested previously that zoonoses are more climate-sensitive than pathogens restricted to humans, due to their wider host and environmental ranges [10]. Climate change is usually not the sole factor causing changes in disease transmission. Changes in the incidence and/or geographical range of CSIs can also arise from interactions between environmental and other factors, e.g. wildlife distribution and changes in land use, that might increase the exposure of local societies and ecosystems. The societal vulnerability may also increase, due to less efficient surveillance and control programmes for CSIs, poor access to veterinary and human healthcare, low education level, inequity and low adaptation to e.g. increasing temperatures. Climate change may increase these and other stressors that affect animal and public health. However, our additional literature search comparing awareness of climate influences in two periods (1997–2006 and 2007–2016) showed that the number of papers studying the effect of climate change on different infections increased significantly (P < 0.01) between the periods.
Characterisation of potential CSIs based on the literature search showed that diseases classified as CSIs are dependent on the ambient temperature, humidity, vegetation cover, surface water or other environmental variables. Arthropod vectors are in general highly impacted by abiotic factors and a changing climate involves changes in temperature and precipitation patterns, which are manifested e.g. in earlier greening and an extended length of vegetation period. Higher temperatures in Northern areas may increase successful overwintering and overall survival of vectors and animal reservoirs, allowing them to expand their distribution range if climate factors have previously been a constraint [2]. High humidity and access to water are crucial for most arthropods, while drought could be detrimental [19].
Leptospirosis was the most dominant disease identified in the food-, feed- and waterborne category. Climate change may alter the habitats and feeding patterns of wildlife species. For domestic animals, new feed crops or changes in feed handling may increase the risk of spread of infectious diseases. Drinking water reservoirs may be contaminated after heavy rain and surface run-off. Flooding and drought may result in water of lower hygienic quality being used.
The number of abstracts on potential CSIs in the soil- and natural water-borne category was limited and, in terms of epidemiology, this is a divergent category of diseases. Spore-forming bacteria, such as B. anthracis and Clostridium spp., may be spread from soil during extreme weather events, such as flooding, landslides and drought [20]. Most abstracts within this category did not focus on climate change and none of the diseases included was classified as climate-sensitive. However, anthrax received much attention in a study by Walsh et al. [21] on anthrax emergence in the warming North, which identified climate as one of several important factors to include in predictive models. Anthrax spores can be resistant to extreme environmental conditions and can survive for decades in soil [22]. When uncovered, the spores can develop into an infective stage, infecting grazing animals. In one recent example due to the thawing tundra, a study based on DNA sequencing and using protein analysis to categorise permafrost-dwelling microorganisms showed that the release of infective spores from old buried animal carcases caused an outbreak of anthrax in Yamal, Russian Federation, that killed approximately 2500 reindeer and caused many human cases, of which one was fatal [23]. Other diseases in this category may be wind-borne and mainly occur following drought, with q-fever being a relevant example.
Only two of the abstracts evaluated, studying pasteurellosis and parapoxvirus (orf), respectively, were considered to belong to the contact transmission category. One reason for this may be that the other four potential CSIs in this category mainly cause problems in reindeer and other ungulates and may not be much studied with respect to influence of climate change. Opportunistic infections are probably also more relevant for animals, especially wildlife. In domesticated animals, management strategies to reduce heat stress or vaccination may mask the effect of climate change on CSIs. However, actions to mitigate the negative effects of feed shortages, such as corralling and supplementary feeding of semi-domesticated reindeer, could pose an increased risk of spread of infection [24]. In our expert discussions these infections were also characterised as potential CSIs, even if the climate change impact is more indirect and not as obvious as for VBD and wildlife-borne diseases.
When wildlife act as a reservoir of a pathogen or are linked in other ways to the epidemiology of a disease, this often intersects with some or all the transmission categories defined here. Wildlife are dependent on climate variables for their geographical distribution, population dynamics, persistence, migration routes etc. [5]. Results on wildlife as intermediate host, vector, amplifier or reservoir category showed that fasciolosis was dominant and was classified as climate-sensitive. Some of the evidence on fasciolosis found in the literature search was from Mexico, in the South. However, a freshwater snail is always involved in the transmission cycle of fasciolosis and thus wet grassland and mild winters most probably increase the risk of its transmission world-wide. Caminade et al. [25] modelled recent and future climate suitability for fasciolosis in Europe and showed that it increased in central and north-western Europe during the 2000s. This simulated trend is consistent with an observed increase in infected ruminants. The simulation results also showed that recent trends are likely to continue in the future in Northern Europe and will most probably extend the season suitable for development of the parasite in the environment [25]. Hantavirus was also important in this category and highly represented in the study area. The literature search yielded no similar support for five other infectious agents: Erysipelothrix rhusiopathiae, Fusobacterium necrophorum, and alphaherpes, gammaherpes and pestivirus.
Our literature search in several databases to identify potential CSIs, using a One Health approach, applying a Northern latitude perspective and assessing potential change in awareness of climate change effects on infections in publications over time, showed that VBD, and in particular TBD, poses an increasing threat for high-latitude regions. This supports findings by McIntyre et al. [10], who studied climate influence on animal and human diseases in Europe. In addition, several ambitious efforts have been made to review the impact of climate change on human diseases [12, 26, 27].
In the present analysis, we considered the fact that the word ‘weather’ was used more often than ‘climate’ in most of the abstracts we evaluated and that long-time weather changes are not always referred to as climate change. Thus, the present study provides an indication of several infectious diseases that are most likely to be CSIs and identifies four infectious diseases as climate-sensitive.
The selection of potential CSIs in the present study was subjective and biases might be present. For example, uncommon diseases, present in only one or a few species with restricted expert knowledge, may lead an infectious disease being favoured by one evaluator or rejected by another. A recent disease outbreak and/or increased attention to a disease in the media or in scientific publications may have contributed to bias in inclusion. The search terms used, the exclusion of publications without an English abstract and trends of interest to obtain research funding for a specific pathogen may also have introduced biases. However, these possible biases were likely mitigated by our stepwise approach, i.e. expert discussions, identification of literature, screening of titles, evaluation of abstracts and evaluation of full papers. Further, when organising potential CSIs into different categories, the most general subtype/serotype of the microorganism of a suggested CSI were discussed. However, some CSIs, represented by different subtypes or serotypes, may differ in epidemiology and could therefore be placed in different categories.
The study was based on the literature representing current knowledge (to October 2017) regarding changes in ecosystems and the impact on disease distribution and provides an indication of infections that may be regarded as CSIs. Yet, climate-affected ecological processes are dynamic, and therefore diseases may fall in or out of the climate-sensitive definition over time.