Quantifying hydroclimatic change impacts on infectious diseases : Signals and geographies from local to global scale

Sammanfattning: Hydroclimatic change has the potential to directly or indirectly increase the occurrence and expand or shift the geographical range of infectious diseases. This may pose particular threats in the Nordic-Arctic Region, where warming is more rapid than in other parts of the world, but the climate sensitivities of various infectious diseases still remain to be investigated in this and other regions. This thesis aims to further our understanding of and predictive capability for the relationships between hydroclimatic change and infectious diseases. To achieve this aim, statistical correlation relationships were analyzed between seven potentially climate-sensitive infectious diseases and a range of hydroclimatic variables across various geographical scales and parts of the Nordic-Arctic Region. The studied diseases were: borreliosis/Lyme disease, tularemia, leptospirosis, Q fever, TBE, Puumala virus infection, and cryptosporidiosis. Hydroclimatic sensitivity has also been investigated through a statistical disease model, site-specifically parameterized at local scale, for the case of tularemia at different Swedish sites (counties) and for different scenarios of future hydroclimatic change. Moreover, for the relatively widespread Lyme disease and cryptosporidiosis, a scoping review approach has been applied to investigate how the complexity of the hydroclimate-disease relationships is considered and quantified in research so far and what key research gaps remain to be bridged.Results identify distinct hydroclimatic variables that are significantly correlated with six of the seven studied human diseases at large spatial scale over the Nordic-Arctic Region. The indicated hydroclimatic disease-driving variables and associated change relationships are to some degree consistent with previous reasoning-based discussions of climate-sensitivity of infectious diseases as increasing threats for humans. Notable exceptions are TBE and leptospirosis, which tend to decrease with increasing regional temperature and precipitation. Borreliosis (Lyme disease) exhibits consistent climate sensitivity at different geographical scales and region parts, considering the whole or either the southern or the northern part of the studied Nordic-Arctic Region. In contrast, tularemia does not exhibit any particular climate sensitivity signal at the large regional scale, even though such sensitivity is evident in local-based statistical disease models. This shows that, in general, investigations at multiple geographical scales and regions, and with different quantitative approaches are needed to obtain a complete picture of hydroclimate-disease relationships. Furthermore, along a latitudinal gradient across Sweden, the likely most realistic medium climate forcing scenario indicates future disease decreases (intermittent or overall) for the relatively southern Swedish counties, and disease increases of considerable or high degree for the intermediate and more northern counties. The projections also show that scenarios of steeper future climate warming do not necessarily lead to a steeper increase in future disease outbreaks and that uncertainties in the disease projections may be large and stem from both the disease models and the climate models. Important research gaps are further identified in research so far on the hydroclimate-disease relationships for Lyme disease and cryptosporidiosis. The gaps regard in particular water-related and socioeconomic factors for Lyme disease, and land-related factors for cryptosporidiosis. For both diseases, climate and other driver-pressure interactions with host and parasite communities are overall understudied. In addition, Asia and Africa emerge as main geographical research gaps for Lyme disease and cryptosporidiosis, respectively. Overall, the consistencies and controversies emerging from the statistical analysis, the uncertainties appearing in the scenario projections, and the research gaps identified by the scoping review in this thesis indicate possible biases in our understanding of hydroclimate-disease relationships and propose relevant directions for future research.