Landscape biodiversity correlates with respiratory health in Australia

Megatrends of urbanisation and reducing contact with natural environments may pose a largely unappreciated risk to human health, particularly in children, through declining normal (healthy) immunomodulatory environmental exposures. On the other hand, building knowledge of connections between environments, biodiversity and human health may offer new integrated ways of addressing global challenges of rising population health costs and declining biodiversity. In this study we are motivated to build insight and provide context and priority for emerging research into potential protective (e.g. immunomodulatory) environmental exposures. We use respiratory health as a test case to explore whether some types and qualities of environment may be more beneficial than others, and how such exposures may compare to known respiratory health influences, via a cross-sectional ecological epidemiology study for the continent of Australia. Using Lasso penalized regression (to interpret key predictors from many candidate variables) and 10-fold cross-validation modelling (to indicate reproducibility and uncertainty), within different socio-geographic settings, our results show surrogate measures of landscape biodiversity correlate with respiratory health, and rank amongst known predictors. A range of possible drivers for this relationship are discussed. Perhaps most novel and interesting of these is the possibility of protective immunomodulatory influence from microbial diversity (suggested by the understudied 'biodiversity hypothesis') and other bioactive agents associated with biodiverse environments. If beneficial influences can be demonstrated from biodiverse environments on immunomodulation and human health, there may be potential to design new cost-effective nature-based health intervention programs to reduce the risk of immune-related disease at a population level. Our approach and findings are also likely to have use in the evaluation of environment and health associations elsewhere. (C) 2017 Elsevier Ltd. All rights reserved.