By Andrew Wu
Climate change has many implications for public health, particularly on the transmission of infectious diseases. Changes in humidity can lead to an increased risk of illnesses that spread through bodily fluids. Vectors such as mosquitoes can become more abundant and affect larger regions. Natural disasters can destroy healthcare infrastructure, alter the immunity of a population, and increase exposure to water-borne diseases. Although there are many factors that modulate infectious disease dynamics, it is crucial that researchers pinpoint associations between the spread of maladies and environmental changes, as they become more drastic and prominent in our lifetimes. A better understanding can lead to more precise models, which can enhance the accuracy of predictions and lead to more effective healthcare. Recently, Professor Metcalf of the Ecology and Evolutionary Biology and Public Affairs departments published a review paper that thoroughly analyzes techniques that investigate the links between climate change and infectious diseases.
There is a multitude of challenges in identifying associations between climate factors and infectious disease dynamics. One of the main issues is the lack of useful data. For example, the information for many pathogens focuses on the incidence of human cases instead of the prevalence of infected vectors and hosts with mild to no symptoms. While both types of data are equally important, more information is needed from the latter for the bigger picture. The latter also requires biological knowledge, which may only be obtained through experimentation. This presents another challenge, since there are few diseases with effective animal models or vectors that can be easily studied. Data is also rarely collected over a sufficiently long timespan in order to reflect shifts in climate, and sometimes the information gathered is only available over a large time interval (e.g. monthly, annual), which will not account for details such as a spike in transmission caused by one day of strong rainfall or a natural disaster. In general, it is important to identify the short-term effects of seasonal events and the long-term effects of climate change. Other factors must be considered as well, including drug resistance and human behavior (e.g. population growth, mobility, public health efforts), which can obscure climate influences on disease transmission. Ultimately, researchers must distinguish between coinciding and corresponding phenomena in order to establish a clear association between climate change and infectious disease dynamics.
Researchers have used techniques in order to build better predictive models for the transmission of various diseases and have a more thorough understanding of the foundational mechanisms. There are two main types of models: time series methods and spatial methods. The former involves two approaches. On one hand, there are traditional methods that utilize statistics more heavily (e.g. generalized linear models, autoregressive integrated moving average (ARIMA) models). Essentially, ARIMA models use the weighted sum of past values in order to estimate future values. They offer more reliable forecasts for shorter time intervals and have been used to predict areas of high risk to dengue due to climate characteristics in Sri Lanka. In addition, statistical approaches were used to observe seasonal variability and Streptococcus pyogenes infections in Iceland from 1975 to 2010. On the other hand, dynamic models focus more on mechanisms and offer more flexibility to account for the long-term changes and variables in climate. They are used to examine various diseases, including the relationship between the onset of influenza and humidity in the United States; the association between rainfall and cholera in Haiti; and the connection between temperature, mosquito population, and dengue in Madeira, Portugal. Spatial methods involve static versus dynamic risk maps, which depict the risk of infection on a regional, national, or global scale by generating a map based on data points from specific locations. Global risk maps were used to pinpoint the role of the 2015 El Niño climate phenomenon with the spread of Zika virus in South America.
Previous models have been successful in predicting the spread of certain illnesses, but there is much work to be done moving forward. As mentioned above, different types of models have their own shortcomings. Statistical do not offer the same long-term reliability of dynamic, mechanistic models. However, the latter does not always provide the short-term accuracy of traditional methods. In addition, dynamic models need to be integrated with disease incidence data, so that they can be honed in order to reflect the observed reality. There continue to be gaps in the information available, which can be remedied using experimental studies. Furthermore, climate and disease models must be combined in order to verify that they hold true. Regardless, it is without a doubt that researchers will have to continue to adjust their models. Despite the challenges that lay ahead, this effort has already had many successes. Models have been used to observe that warmer locations are spreading dengue in Southeast Asia, give accurate real-time predictions for influenza in US cities, and suggest possible malaria spread in the future under climate change. These examples show that better models can provide governments and public health institutions with more precise warnings in order to bolster preparation, which will lead to a healthier society in an ever-changing climate.
Image found on (https://www.cdc.gov/climateandhealth/images/climate_change_health_impacts600w.jpg)
Image found on (https://www.sei-international.org/mediamanager/images/Events/SEI-2015-Events-AdaptationToClimateChange-Flickr-7350728738_a73becd505_o-550x.jpg)