The recent death of Paul Epstein, a physician who advocated for epidemiologists to consider the health impact of climate change, has renewed conversations about our future scientific path to understanding highly complex interactions between global weather and human disease. Ever since Epstein’s 1999 commentary in the journal Science, which argued that global warming would enhance the emergence of infectious diseases, epidemiologists have been avidly working to determine how to mitigate the negative implications of climate change on human health—from predicting outbreaks, to cleaning up superfunds, to counter-acting the marine pollution that followed the recent oil spill in the Gulf of Mexico. One particularly interesting set of studies have identified climate change as a major factor in food safety—affecting the risk of zoonotic diseases, mycotoxin contamination, biotoxins in fishery products and environmental contaminants. In this week’s blog, we look at the food safety implications of climate change, from algal blooms to new strategies for surveillance.
Relationships between climate change and human health
We’ve transitioned from using the term “global warming” to employing the term “climate change”, in part because increased average global temperature isn’t the only pattern being observed; other effects of climate change include stronger storm systems, increased frequency of heavy precipitation and extended dry periods. The contraction of the Greenland ice sheet is anticipated to cause a rise in sea-levels, and it has been argued that the effects of these phenomena will be inequitably distributed, since most of the actions causing climate change originate from the developed world, but the less developed world is likely to bear the brunt of the public health burden: declines in water quality and quantity, particularly in already dry regions; infectious disease transmission extending to newly-warmed areas; and hurricanes, floods and mudslides affecting the poorest area, leading to stagnant water-borne diseases like cholera.
Climate change is likely to have other specific impacts on food production. By increasing the populations of pests in currently temperate regions, climate change is likely to lower the yield of major crops and induce storms that damage those crops. Changes to the availability of feed, as well as heat stress, are also anticipated to affect animal production, and migration of fish away from coastal regions (affecting food availability) have also been observed.
Back in 800 BC, Homer catalogued the impact of ocean pollution in The Odyssey, writing about the death of sailors from contaminated fish. Today, ocean-induced poisoning is often related to consumption of contaminated seafood, particularly following harmful algal blooms (HABs)—accumulations of algae in the water (sometimes associated with a “red tide”), which have been found to produce neurotoxins and contaminate fish. Over 60,000 cases of poisoning by exposure to HABs are reported in the US each year. Some of the more dramatic cases involve paralytic shellfish poisoning, but chronic liver disease and other more indolent pathologies also result from HABs. Climate change and ocean warming appears to increase eutrophication (nutrient loading) causing phytoplankton growth, increasing the frequencies of HABs and especially the build-up of toxic species of algae—in addition to contributing to higher mercury uptake among fish and accumulation of toxic Vibrio bacterial species.
The data on food safety and climate change
Early evidence for the relationship between climate and food safety came from analyses showing that foodborne and diarrheal diseases exhibited a seasonal incidence pattern, increasing with average temperatures and after severe weather events. More recently, epidemiologists have noted increases in salmonella and campylobacter infections after weeks of elevated ambient temperature. Rates of salmonella in Australia specifically vary with increasing average yearly temperature, and higher humidity has also correlated with decreased hospitalization rates for children diagnosed with food-borne rotavirus (after correcting for other confounding factors), likely because survival of the virus is favoured at lower humidity. El Nino-associated rises in cholera have been documented in both Peru and Bangladesh, as have increases in diarrhoeal disease in Peruvians. HABs and related phytoplankton proliferation seems to increase environmental pH, giving Vibrio cholerae a competitive advantage over other marine bacteria since it thrives at higher pH and promotes attachment of V. cholerae cells to zooplankton (particularly copepods) which protect V. cholerae cells from external stresses.
Recently, zoonotic diseases—those transferred from animals to humans—have been observed to contaminate food and water sources in the context of climate change. This appears to result from increased susceptibility of animals to disease, increased range or abundance of vectors and animal reservoirs for disease pathogens to hitch-hike on, and prolonged transmission cycles among vectors. Susceptibility of animals to disease seems to be occurring in the context of heat waves; high water temperatures, for example, inhibit the functioning of oysters’ immune systems, which precipitated a series of outbreaks of V. paparahaemolyitic among humans who consumed oysters from northern waters (e.g., Alaska) between 1997 and 2004. Cycles of drought followed by heavy rainfall have also been found to enhance breeding sites for midge and mosquito vectors, which are associated with outbreaks of livestock diseases (similar patterns have been observed with the expansion of tick breeding territories). In 2000, global temperature changes were associated with strains of Rift Valley Fever (that probably originated in east Africa) escaping from Africa for the first time and infecting the Arabian Peninsula, an area well connected to Europe by a ‘ruminant street’ (rodents hitch-hiking on trade vessels). Changes to climate were thought to facilitate the survival of rodents and the overlap of rodent populations to a degree not previously thought possible.
Finally, a number of concerns have been raised from mycotoxins: a group of highly toxic chemical substances (like aflatoxin) that are produced by toxigenic moulds like Aspergillus that commonly grow on a number of crops when temperatures rise. In recent years, outbreaks of acute aflatoxicosis have been reported for the first time; 125 deaths occurred in Kenya for example, out of 317 reported cases resulting from consumption on aflatoxin contaminated maize in 2004, followed by repeat events in 2005 and 2006. Benin and Togo have now reported similar occurrences in the context of annual temperature increases. Aspergillus outbreaks in Europe and the United States have now started to occur for the first time, following extended droughts.
Approaches to prevention
Mathematical models have recently been used to predict when HABs or other sudden outbreaks of harmful diseases will occur. Inferences can be made based on patterns from existing data, such as by studying disease trends in the presence and absence of El Nino events. For example, one group recently used time series analysis along with epidemiological data to predict disease outbreaks caused by three foodborne pathogens (Salmonella, Campylobacter, and E. coli), warning inspectors about the times they needed to be particularly vigilant or have extra human-power available for testing food samples. Another group used remote sensing data to indirectly measure V. cholera behaviour as a function of ocean temperature and surface height, providing a means by which to predict conditions conducive to pandemic disease. This resulted in keeping sensors along coastal areas to detect early HABs and warn fishing regulators.
A second approach to addressing the problem is through molecular ecology, in which nucleic acid sequence comparisons and other genomics-based approaches are used to characterize complex microbial communities, identifying which may be pathogenic or produce important toxins. These methods are applicable to the study of microbial evolution during periods of increasing temperature, including virulence factor acquisition and changes in gene expression (potentially affecting growth of algae or production of toxins) due to environmental exposures. When combined with remote sensing and geographic information systems, it should be possible to use this information to model the distribution and spread of different types of pathogens as a function of temperature or other climate change variables.
Perhaps most important of all is proactive surveillance through direct detection of pathogens in foods and the environment: old school, door-to-door epidemiology. This requires a lot of human effort, but can be aided with molecular methods. One problem is that common molecular laboratory tools like PCR can’t tell between inactive and active pathogens in a food sample, while activity-based assays like enzyme immunoassays aren’t very specific and can produce a lot of false positive results. An ideal method would provide rapid field-based testing of dangerous pathogens in water, crops or animal products. While this technology is not yet available, a surrogate is used such as the detection of fecal organisms and E. coli to predict fecal contamination of water. Unfortunately, these surrogates often do not correspond well to cholera or other pathogenic bacterial contamination levels.
The other problem with developing a strategy to detect environmental contaminants is not technical, but political. As the UN’s Food and Agriculture Organization (FAO) recently specified, building capacity to detect contaminated food is often counter-productive from an economic standpoint: the surveying country would not want to face the economic trouble of declaring that their exported food is potentially hazardous. Furthermore, as food-borne illness or unsafe foods typically affect the poorest groups, building safe food stocks is often not considered by powerful politicians who rarely engage with the lower class.
The FAO and others have therefore engaged in a mutual-risk/mutual-surveillance scheme in which international modelling and monitoring of HABs and other risks to food is publicized widely, just as the WHO publicizes infectious disease outbreaks, in the hope that such attention will force politicians to engage in risk mitigation. Whether or not this “Emergency Prevention System for Transboundary Animals and Plant Pests and Diseases” (EMPRESS) program will work (and whether anyone can remember what the acronym stands for) has yet to be tested in a real-world emergency. But increased attention to the problem from epidemiologists like Epstein is likely to at least maintain active surveillance and build initial capacity.