Societies depend heavily on the ocean for food, transportation, recreation, waste disposal, and minerals. Clearly, some of these uses are in conflict, and four of these five major uses provide a vehicle for transmission of disease agents to humans. The primary source of marine-borne illness is seafood (Czachor, 1992; Lipp and Rose, 1997). Currently seafood is in high demand, both in industrialized and emerging nations, and it continues to be a source of infectious diseases in humans throughout the world (Garrett et al., 1997; IOM, 1991). In the United States, the demand for seafood is so high that fish and fish products were the third leading contributor to the U.S. international trade deficit, reaching a staggering $4.2 billion in 1997 (McCarthy, 1996; NMFS, 1998).
Transportation has also contributed to waterborne infectious diseases through the consumption of contaminated water and/or seafood by ship passengers and crew. There is also evidence that ships are responsible for the dissemination of exotic species, including human pathogens, through the discharge of bilge water into coastal waters (NRC, 1996). Recreation is another source of exposure through ingestion of seawater or contact of skin and/or mucous membranes with seawater. Finally, disposal of human wastes in the ocean can lead to diseases of plants and animals, including humans.
In the final decade of the 20th century it appears that waterborne diseases of humans are as prevalent as they were at the start of century. For example, waterborne illnesses continue to be a major killer of children throughout the world
(ICDDR,B, 1998). It is estimated that 60% of the world's population lives in coastal zones, the area within a few kilometers of the shoreline. Disease incidence is increasing worldwide, promoted by both natural phenomena such as El Niño and human activities, including sewage disposal. Ancient diseases like cholera still cause epidemics (Plate V) while new agents of disease (e.g., hepatitis E and Vibrio vulnificus) continue to be discovered. Although the pathogenesis of diseases such as cholera and dysentery is well understood, the cause of outbreaks of these diseases is unresolved. Experts still debate how cholera epidemics arise despite a detailed understanding of the genetics, chemical structure, and mode of action of the cholera toxin. Cholera is known to reside in human hosts and spread by the oral fecal route, however, epidemics may be seeded by vibrios that reside in estuaries and other saline waters and infect people through contaminated drinking water and seafood. The molecular structure of recently discovered disease agents has confirmed their existence on Earth for thousands of years. The apparent emergence of these pathogens could be the result of anthropogenic influences or may reflect more sensitive modern detection technologies. Whatever the cause, it is clear that infectious diseases, including waterborne diseases conveyed by the ocean, still plague humankind.
The Agents of Waterborne Disease
The principal agents of diseases that derive from seawater and seafood are viruses and bacteria. Most disease appears to result from ingestion of contaminated seafood (IOM, 1991), although accidental ingestion of seawater (e.g., recreational exposure or contamination of potable water with seawater) is another important route of infection. Some agents enter the human body through wounds (e.g., puncture wounds from sea urchins and cuts from fishing gear) othersLeptospiraalthough rarely contracted from seawater, enter through broken skin, or penetrate mucous membranes, and a fewavian schistosomespenetrate unbroken skin.
Viruses are obligate parasites and require living cells for reproduction. Certain viruses, however, can survive in seawater for long periods of time (e.g., hepatitis A, poliovirus) and are concentrated by marine bivalve mollusks, such as oysters and clams. Human waterborne viral infections result from contamination of seawater or seafood by sewage.
Unlike viruses, most of the pathogenic bacteria do not require human hosts for replication. Indeed, some are naturally occurring in estuaries and the coastal ocean, and can grow on, or within, many animals, ranging from zooplankton to fish, and on marine plants, ranging from phytoplankton to macrophytes. In general, bacteria that cause seawater- and seafood-borne diseases in humans come from two different sources. The foreign or allochthonous bacteria come from humans and other animals by means of fecal contamination: sewage outfalls, septic tanks, and land surface runoff. One exception to this generalization for
allochthonous bacteria is Leptospira interrogans; it lives in the urinary tract of mammals, especially rodents, and enters water by means of contaminated urine. The indigenous or autochthonous bacteria are usually commensals of marine plants and animals, although some may be free-living in the water. For example, the vibrios are indigenous to estuaries and the ocean, and several species are human pathogens. Vibrio cholerae causes human cholera and V. parahaemolyticus causes gastroenteritis and wound infections. Because of the popularity of seafood in Japan, V. parahaemolyticus is one of the most common causes of gastroenteritis there (IASR, 1996). Examples of the more common microorganisms that cause human disease and are conveyed by seawater are listed in Table 2-1. Detailed discussion of these common microbial agents of disease can be found in most textbooks of microbiology, including Howard et al. (1994). Grimes (1991) has reviewed the literature on estuarine bacteria capable of causing human disease.
In addition to the viruses and bacteria, there are a few waterborne animal parasites and fungi that cause human disease. The best known are the avian schistosomes that cause swimmer's itch, the nematode that causes anisakiasis (Deardorff and Overstreet, 1991), and the yeast that causes candidiasis. Outbreaks of infection by these agents are relatively rare (toxic algae are also microbial agents of human disease, and will be discussed in Chapter 3 of this report). Stings inflicted by anemones, jellys, and stingrays associated with recreational use of coastal waters provides sites for infection. Such infections usually result from autochthonous bacteria such as the vibrios (Thomas and Scott, 1997).
Detection and Prevention
Classic detection methods for microbial agents of disease have relied on microscopy and cultivation. Microscopy is useful only for the larger organisms that can be identified by their morphology. Fluorescence microscopy increases the options for visual identification by using fluorescent dyes coupled to bacteria- or virus-specific antibodies or gene probes. The traditional method for detecting viruses and bacteria is cultivation. Suspect samples of water and seafood are inoculated into nutritive broths or on nutritive solid media (bacteria), or tissue cultures (viruses). Positive cultures allow for the identification of the agents. In some tests, samples are injected into animals, which are then monitored for symptoms of disease.
Although cultivation is often successful in isolating pathogens, it is now recognized that many bacteria remain undetected because they are in a viable but nonculturable state (Grimes et al., 1986; Roszak and Colwell, 1987; Xu et al., 1982). They can be visualized by microscopy, but they cannot be grown using currently available media and protocols. Most of the bacteria listed in Table 2-1 are capable of entering a nonculturable phase (Grimes et al., 1986; Roszak and Colwell, 1987), and no methods for their cultivation from this phase have yet been formulated. Hence, testing water and seafood for the presence of pathogenic bacteria by using culture techniques can be misleading.
Detection and counting of specific bacteria, historically referred to as fecal indicator bacteria, have comprised the generally accepted method for estimating the public health safety of both drinking and recreational water and seafood (Dufour, 1984). This technique relies on the culture of coliforms, a term used to include several genera and species, including Escherichia coli, Enterobacter aerogenes, Klebsiella pneumoniae, and even some Salmonella serotypes, thought to derive from the colon of warm blooded animals. Originally, this group of bacteria was employed to estimate the extent of fecal contamination of water samples and, hence, the potential of enteric disease (Dufour, 1984). However, it was realized as early as 1900 that some coliforms are frequently associated with natural bodies of water including estuaries. Klebsiella pneumoniae, for example, is commonly associated with plants and can wash into estuaries and the coastal ocean during heavy rains. If present, these non-sewage-related coliforms will
cause falsely high estimates of pollution, making the use of total coliforms as an indicator of enteric disease questionable and sometimes misleading. Although the introduction of the fecal coliform count improved the specificity of the test, false positives from other non-fecal bacteria remain a problem. Today, it is accepted that the fecal coliforms provide public health practitioners with a useful indicator; however, it is also widely recognized that the fecal coliform test has some serious disadvantages:
Finally, it should be noted that certain strains of E. coli are themselves pathogenic for humans (e.g., E. coli 0157:H7), and these toxin-producing strains can survive in seawater in the viable but nonculturable state for several days to weeks (Grimes and Colwell, 1986). For example, a recreational outbreak of E. coli disease occurred in a water park in June, 1998 (CDC, 1998a), infecting 26 children and causing one death. Clearly, E. coli is more than an indicator of fecal pollution; it may be a preferred indicator of contamination for freshwater and possibly low salinity water. Certain pathogenic strains of E. coli are themselves candidates for surveillance.
The fecal coliform test continues to be the standard used for shellfish and waters where shellfish are grown, as established by the National Shellfish Sanitation Program (NSSP; FDA, 1996). Although numerous studies have reported that fecal coliforms do not correlate with any of the current diseases that can be contracted
from eating raw or undercooked oysters (Goyal et al., 1979; Koh et al., 1994), there is a reluctance for the NSSP to abandon them as its index of public health safety.
Some investigators have suggested that enterococci may be the preferred indicator, but these microorganisms have some of the same limitations as the fecal coliforms (Koh et al., 1994). Because of these limitations and reliance on the existing database for fecal coliforms, the enterococci have not been universally accepted by state, county, and local regulatory agencies as a substitute indicator group, even though some states have adopted the new E. coli and enterococcus standards. In addition, the presence of enterococci or E. coli does not always correlate with the presence of viruses.
New indicator systems are being proposed, some of which are based on molecular genetic methods. The coliphages, viruses attacking the coliform bacteria, have been suggested as an indicator of risk of infection with human enteric viruses, including hepatitis A virus. They have a quantifiable association with specific viral pathogens representing an improvement over fecal coliforms (Paul et al., 1997). Presence of Clostridium perfringens spores may be helpful in distinguishing long-term accumulation or movement of sewage in aquatic habitats. This is because C. perfringens is fecal specific and the spores can survive for many years (Emerson and Cabelli, 1982). Of greatest interest, however, are the gene probes now being used in clinical laboratories to detect pathogenic microorganisms. Gene probes can be highly specific, capable of detecting genetic sequences common to, or evolutionarily conserved in, pathogens of concern. For example, gene probes can be used in environmental testing to detect, directly identify, and quantitate pathogens such as Salmonella thus obviating the need for culturing (Brasher et al., 1998). Also, amplified nucleic acid sequences, using the polymerase chain reaction (PCR), allows detection of very small numbers of bacteria and viruses in a given sample. For example, Brasher et al. (1998) have developed a series of gene probes that show promise for the direct detection of enteric pathogens in seawater and seafood. Monoclonal antibodies also have been shown to be useful for direct detection of pathogens in estuarine and seawater samples, especially when coupled with fluorescent dyes (Huq et al., 1990). Finally, the use of immunoglobulin A as a human-specific indicator of fecal pollution has received some recent interest (Middlebrooks, 1993). It is predicted that the new biotechnological methods will eventually replace the cumbersome and inaccurate fecal indicator tests for public health assessments of the safety of water and shellfish. For a recent review of detecting viruses in the environment, the reader is referred to Metcalf et al. (1995).
Coastal Oceans, Estuaries, and Waterborne Disease Transmission
Allochthonous materials enter estuaries and the coastal ocean by four routes: point sources, nonpoint sources, direct disposal or discharge, and atmospheric
fallout or rainout. Point sources include sewage outfalls and industrial wastewater outfalls. Nonpoint sources include storm runoffs, storm sewer outfalls, and rivers. Both point and nonpoint sources of pollution have the potential to contaminate receiving waters with human pathogens. In past years, disposal of sewage sludge contributed pathogens to both the water column and marine sediments. Direct disposal of wastes into the ocean is now largely prohibited as a result of the London Dumping Treaty (Zeppetello, 1985) and the Ocean Dumping Act of 1988, and, therefore, refuse discharge is no longer a major source of human pathogens in the ocean. A recently recognized source of human pathogens is bilge and ship ballast discharge. Ships that take on ballast water in foreign ports can inadventently take on water contaminated with pathogens and then discharge those pathogens into previously uncontaminated waters. The National Research Council recently published a report on the introduction of non-native species by ballast water (NRC, 1996), and McCarthy and her colleagues have studied this problem in the Gulf of Mexico (McCarthy and Gaines, 1992; McCarthy and Khambaty, 1994; McCarthy, 1996). The scope of this problem is not yet clear, but ship ballast discharge has the potential to spread microorganisms throughout the world's coastal ocean. Less clear are contamination problems from atmospheric fallout or rainout, but a recent publication suggests that pathogens can be spread by air (Rosas et al., 1997). Contributions due to atmospheric fallout or rainout occur, but are not well understood (Rosas et al., 1997).
Point sources of allochthonous materials continue to be an important source of pathogens in the estuaries and the coastal ocean of the United States. Sewage treatment plants in the United States discharge approximately 2.3 trillion gallons of effluent into marine waters annually, and more than 2.8 billion gallons of industrial wastewater are released daily (NOAA, 1998c). Even though much of this effluent is receiving some type of treatment prior to discharge, it is still contributing pathogens, nutrients, and toxic chemicals to the coastal ocean of the United States. While point sources therefore continue to cause degradation of coastal areas, it is now generally believed that nonpoint sources are equally if not more important. Nationwide, it has been estimated that indirect loadings account for more than half of the suspended solids, nutrients, fecal coliforms, and metals entering coastal waters annually (NOAA, 1998a). The two largest sources of nonpoint pollutants are agricultural runoff and storm sewer discharge.
It is estimated that over 9.1 billion gallons of domestic sewage and industrial wastewater enter the coastal ocean of the United States each day (NOAA, 1998a). When combined with accidental spills (e.g., oil spills) and nonpoint source runoff, the net result is a vast and complex mixture of chemicals, both organic and inorganic, entering the coastal ocean of the United States. These chemicals primarily cause two alterations in coastal ocean ecosystems. Some serve as growth stimulating nutrients, either directly as sources of food for marine plants, animals, and microorganisms or indirectly as essential growth factors (e.g., vitamins). Other chemicals are toxic to living organisms, causing alterations in
growth (e.g., stunting and neoplasms) or death. In some cases, chemicals that are toxic to some forms of life can actually serve as growth stimulating nutrients for others, especially microorganisms. A good example of this is the ability of carcinogenic polycyclic aromatic hydrocarbons (PAHs) to serve as a growth medium for certain marine vibrios, including pathogenic vibrios (West et al., 1984). Another example is the shift in bacterial communities that was described in the mid-Atlantic Ocean by Grimes et al. (1984). Thus, it is possible that certain toxic chemicals entering the coastal ocean of the world have multiple effects. Some of these chemicals could be making fish and marine mammals more susceptible to disease by altering their resistance (immune system), while at the same time stimulating the growth of agents capable of causing the disease. It is worth noting that the vibrios are capable of degrading a wide variety of organic compounds, including aliphatic and aromatic hydrocarbons, and this generalization includes the pathogenic vibrios. However, proper experiments have not been done to establish linkages between toxic chemicals, declining fisheries, human disease, and densities of pathogenic vibrios.
Most of the waterborne and seafood-borne diseases throughout the world are caused by viruses. The major agents of waterborne viral disease are listed in Table 2-1, and they include several well-known groups. The agents in these groups are all RNA (Ribonucleic Acid) viruses, and most have not been cultured. Viral hepatitis is now known to be caused by two related viruses, hepatitis A and E. These viruses are very resistant to environmental extremes, including toxic chemicals, and they are concentrated from water by filter-feeding shellfish, especially oysters. The human caliciviruses include the Norwalk virus and all cause gastroenteritis. It is estimated that most shellfish-borne and waterborne disease in the United States is caused by the Norwalk virus (GAO, 1984). None of the caliciviruses can yet be cultured and they are detected by antibody-based tests (e.g., Enzyme-Linked Immunosorbent Assay [ELISA]) and RT-PCR (reverse transcriptase PCR). The NRC Committee on Evaluation of the Safety of Fishery Products reviewed the incidence of Norwalk and Norwalk-like agents (IOM, 1991). The rotaviruses and astroviruses are responsible for widespread diarrheal disease, especially in children. It is estimated that rotaviruses kill 870,000 children in the world annually (U.S. Department of Health and Human Services, 1998). As described for the caliciviruses, their detection is dependent on RT-PCR assays and immunologic procedures. Recently, a rotavirus immunization procedure was approved for children. Finally, the enteroviruses represent a large and diverse group of viruses that include the poliovirus. Enteroviruses cause a variety of clinical syndromes, including aseptic meningitis, paralysis, myocarditis, rash, pneumonia, fever, and undifferentiated febrile illness. The only common link among these viruses is their mode of transmission (fecal-oral route, which can include water and food as intermediaries) and their nucleic acid content is RNA. It is doubtful that global climate change events will have any significant direct effect on these viruses, because they are not conveyed by vectors and are incapable
of growing in ocean habitats. However, changing demographics and anthropogenic pressures (e.g., sewage contamination) very definitely contribute to the epidemiology of these diseases. Avian schistosomes are a problem for recreational users of coastal waters, causing what is known as swimmer's itch or clam digger's itch. The cercariae (free-swimming larvae) of these schistosomes penetrate human skin; but humans are not hosts for the parasite and the cercariae die in the skin causing mild to severe itching due to an allergic reaction to the foreign protein. Human anisakiasis is caused by a nematode that infects humans via raw or inadequately prepared seafood (Deardorff and Overstreet, 1991). The incidence of this parasitic disease is very low in the United States, but may be on the increase because of the current popularity of raw fish as a delicacy (e.g., sushi and sashimi). In Japan, over 3,000 cases of anisakiasis are reported annually (IOM, 1991). It is difficult to predict how global climate change events would affect these two parasites. It is possible that increased water temperatures could disrupt life cycles of these parasites by changing the range of their hosts (e.g., snails, birds, fish, marine mammals) and, hence, accessibility to humans. Demographics (e.g., continued population shifts to the coast) and cultural food preferences (e.g., consumption of raw fish) could also have a definite effect.
Global Climate Change and Infectious Disease
One of the first documented associations of a human pathogen with an estuarine animal subject to climate-induced fluctuations was the Vibrio parahaemolyticus-zooplankton relationship described by Kaneko and Colwell (1973) in the Chesapeake Bay. They determined that V. parahaemolyticus overwintered in Chesapeake Bay sediments and entered the water column when water temperatures reached 14 ± 1 ¹C. Upon entering the water column, the vibrios became associated with zooplankton, primarily on the surfaces of copepods. The abundance of the vibrios increased in direct proportion to the growth of the copepod population. Kaneko and Colwell (1973) also demonstrated a direct relationship between water temperature, numbers of zooplankton, and V. parahaemolyticus. Over 10 years later, Watkins and Cabelli (1985) observed a similar relationship between V. parahaemolyticus and zooplankton in Narragansett Bay. In addition, they demonstrated that fecal pollution had an indirect effect on V. parahaemolyticus densities. Nutrients associated with the wastewater stimulated the growth of phytoplankton boosting zooplankton populations, which in turn supported greater densities of the vibrios. Clearly, coastal ocean nutrient enrichment can increase the prevalence of pathogenic vibrios, indicating that climatic events that change the abundance of plankton also have the potential to affect the spread of disease. It is interesting to note that elevated water temperatures have been implicated in the largest reported outbreak in North America of shellfish-borne V. parahaemolyticus. During July and August, 1997, 209 persons became ill and one person died from consuming raw oysters harvested from California, Oregon,
and Washington in the United States and British Columbia in Canada. All 209 infections were culture-confirmed as V. parahaemolyticus (CDC, 1998b). In early summer 1998, the Food and Drug Administration warned consumers not to eat raw oysters from Galveston Bay, Texas, because they might contain harmful levels of V. parahaemolyticus (FDA, 1998). By the end of July, 368 persons had become ill from consuming raw oysters harvested from Galveston Bay, Texas; so far V. parahaemolyticus has been confirmed in 66 of those cases (AP, 1998b).
It has become clear that Vibrio vulnificus also has a preference for warmer temperatures, making it a candidate for the list of pathogens that could be influenced by climate-induced increases in water temperature. Motes et al. (1998) investigated the temperature and salinity parameters of waters where oysters were linked to seafood-borne V. vulnificus infections, and found that abundance of this pathogen was directly correlated with water temperature. Numbers of V. vulnificus increased with water temperatures up to 26 ¹C and were constant at higher temperatures. Salinity also played a factor and high V. vulnificus levels were associated with intermediate salinities (5 to 25 ppt1).
Colwell (1996) noted that coincident with an outbreak of cholera in Peru, an El Niño event that began in 1990 had warmed the nutrient- and phytoplanktonrich waters off the Peruvian coast. The Peruvian outbreak, which began in January 1991, quickly spread to most of the neighboring countries in South America (Mata, 1994). In 3 weeks, the epidemic had covered over 2,000 km of coastal Peru and caused 30,000 cases of cholera; it claimed 114 lives in the first 7 days. Since V. cholerae, like V. parahaemolyticus, associates with planktonic copepods (Colwell, 1996), it can be hypothesized that the South American cholera outbreak resulted from increased levels of V. cholerae that grew in response to the warmer water and nutrients in terrestrial runoff from increased rainfall. Although this conclusion was not reached by direct experimentation, the V. parahaemolyticus model indicates that this proposed link is worthy of further scientific investigation.
More recently, Colwell noted a relationship between sea surface temperature in the Bay of Bengal and cholera cases reported in Bangladesh (Colwell, 1996). A plot of the percent of diarrheal disease caused by V. cholerae versus seasonal sea surface temperature data obtained by remote sensing, revealed a correlation between the cholera case data and increasing water temperature. As with the Peruvian hypothesis, direct experiments to relate the two observations were not performed. However, the data are compatible with the ecology of the vibrios, and should be followed up with carefully planned interdisciplinary studies including epidemiology, meteorology, and oceanography. Plate VI shows that sea surface temperature was elevated from December through April off the western coast of South America during the 1997–98 El Niño. Sea surface height (Plate VII), however,
1 ppt = parts per thousand.
Plate Section has been moved to the back of the book.
peaked during December, and corresponds with the seasonal name for this phenomenon (El Niño refers to the birth of Jesus). These data could be collected and coordinated with water sampling for zooplankton, V. cholerae, nutrients, and chlorophyll, and with epidemiology for cholera incidence.
Storms fueled by the 1997–1998 El Niño increased urban runoff, saturated septic tank leach fields thereby causing overflows, and caused a weakened sewage main to break and spill 120 million gallons of raw sewage into Santa Monica Bay. As a result, public health officials in California closed beaches on 50 days during the period May 1997 through April 1998; this was more closure days than in the three previous years combined for Santa Monica Bay (Alamillo and Gold, 1998).
The ocean influences climate and weather (see Chapter 1) and hence indirectly affect infectious disease agents that are sensitive to changes in temperature and/or rainfall (Pearce et al., 1995). Global warming may affect the incidence of certain protozoal, bacterial, and viral diseases, because increased temperatures tend to change the geographic distribution and reproductive success of many vectors that carry these diseases. For example, higher temperatures may increase the range of some species, but it may also shorten insect survival hence decreasing the transmission of some vector-borne diseases. Indirect effects of climate change may have an impact on human health, but accurate assessment of the risk requires a better understanding of how vectors and their pathogens respond to changes in weather patterns. Indications of how global climate change can affect the spread of vector-borne infectious diseases may be discerned from studying the effects of periodic weather events such as El Niño/Southern Oscillation (ENSO). Weather events such as ENSO bring extremes of both rainfall and drought and change the capability of disease vectors to spread their pathogens to humans. Some of these changes appear to be due to an increase or decrease in habitat, but the spread of vector-borne diseases is also modified by the effects of these disasters on the water supply and public hygiene. Specific effects of ENSO on several of the vector-borne diseases are described below.
Major Disease Agents and their Vectors
Malaria: In the aftermath of a severe marine weather event such as a hurricane or monsoon, there is frequently a higher risk of malaria from an increase in standing water breeding habitat and in some cases from a disruption of malaria control measures. Malaria remains one of the most prevalent parasitic diseases in the tropics. The disease agents are the protozoa, belonging to the genus Plasmodium,
a parasite transmitted by the anopheline mosquito. Worldwide prevalence of the disease is estimated to be 300–500 million clinical cases per year, with 90% occurring in sub-Saharan Africa and most of the remainder in India, Brazil, Sri Lanka, Viet Nam, Colombia, and the Solomon Islands (WHO, 1996). Paradoxically, an increase in malaria transmission may be associated with either drought or rainfall as seen in the Indian subcontinent before the introduction of residual insecticides. In the arid Punjab, rainfall improved the breeding and survival of mosquitoes. However, in the high rainfall area of Sri Lanka, lack of rainfall from the failure of the monsoons led to reduced flow of rivers and the formation of standing pools of water that provided favorable breeding conditions (Bouma and van der Kaay, 1996). Drought associated with the failure of the south-west monsoons are twice as frequent in the year following an El Niño (Dilley and Heyman, 1995) and correlate with past malaria epidemics. Improved predictions of malaria transmission may be possible through the application of hydrological models to estimate available surface water for mosquito breeding (Patz et al., 1998a).
In Venezuela and Colombia the incidence of malaria morbidity and mortality increased substantially (35–37%) in the years following an El Niño. In these countries, the increase correlated more strongly with drought in the first year of the El Niño as opposed to rainfall in the second year (Bouma and Dye, 1997 and Bouma et al., 1997a). It is not clear how drought promotes disease in these countries; however, the authors suggest that in the dry year there is a loss of mosquito predators which allows the mosquito population to increase unchecked in the following wet year.
There has been some question as to whether El Niño conditions gave rise to the epidemics of malaria that occurred during 1983 in Bolivia, Ecuador, and Peru. A review of the data reported by each country on malaria shows that the incidence of this disease began to rise in each of these countries in 1983. However, the overall trend from 1970 to 1996 was an increase in the number of cases reported, while in other El Niño years (1971–1972, 1976–1977, 1991–1992) the incidence of malaria seldom increased over that of previous years. It is known that, during this time, national malaria control programs in Latin America switched from a strategy of rigid eradication to flexible control. This alone could have caused the observed increase. Conversely, a good eradication program may have been masking the impact of El Niño in previous El Niño years. Establishment of an association between an El Niño event and outbreaks of malaria will require future research that emphasizes accurate reporting of outbreaks and consideration of the impact of malaria control measures.
Rift Valley Fever: Rift valley fever (RVF) is a vector-borne viral disease transmitted by the Aedes mosquitoes that primarily affects livestock but which also afflicts humans. Incidence of this disease is highly correlated with excessive rainfall, such that high rainfall in an El Niño year may trigger an outbreak of RVF. However, heavy rains are not sufficient to predict an epidemic of RVF. Transmission is also dependent on the presence of both the competent vector
(Aedes mosquito) and the pathogen. In 1997, the eastern African countries of Kenya and Somalia experienced 60–100 times the normal rainfall in some areas. This was followed by an outbreak of 89,000 cases of RVF in northeastern Kenya and southern Somalia, possibly the largest ever reported (WHO, 1998). Conversely, Kenya did not experience an outbreak of RVF during the 1982–83 El Niño, despite heavy rainfall.
Dengue: Dengue and Dengue Hemorrhagic Fever (DHF) are caused by four distinct, but closely related, viruses carried by the Aedes mosquito. This mosquito has become adapted to urban habitats, breeding in manmade collections of rainwater as well as water stored for drinking and washing. The occurrence of this disease has increased dramatically in the past few decades, from 9 countries prior to 1970 to 41 countries by 1995. Currently, both the Aedes mosquito and dengue viruses are endemic to the tropics worldwide. Each year, dependent on the frequency of outbreaks, there are an estimated 100 million cases of dengue fever, and several hundred thousand cases of DHF, a leading cause of death and hospitalization of children in Southeast Asia (Gubler, 1998). As with malaria, it is difficult to demonstrate scientifically that changes in the distribution of dengue are the result of climate variability. PAHO/WHO, in its preliminary study in the Americas, did not find a correlation between national statistics and increased rainfall (PAHO, 1998b). In fact, peaks in reports of dengue did not occur during past El Niños in the Americas.
It is too early in the current El Niño year to evaluate fully the effects on the incidence of dengue; however, 1998 may be the worst year on record for dengue in some areas. In Viet Nam, there were twice as many cases as of June 1998 than in the previous year and 1997 was the worst year since 1991. Similarly, in Brazil there were as many cases of dengue in June as there were for the entire year in 1997, despite the initiation of an eradication program in 1997. Some researchers have proposed models that link potential changes in global climate with changes in the distribution of dengue transmission (Jetten and Focks, 1997; Patz et al., 1998b). However, the overall spread of the disease during the last few decades argues that factors other than weather are responsible for the current increased prevalence of dengue (Gubler, 1998).
There are several other viral diseases that may be affected by changing weather and climate patterns including yellow fever, hantavirus, and viral encephalitis as well as other parasitic diseases such as schistosomiasis and sleeping sickness since the vectors are sensitive to changes in temperature and rainfall (Rogers and Packer, 1993; Stone, 1995).
Evaluating the Impacts of ENSO on Vector-borne Disease
Future research should aim to improve the quality of the indicators and data on disease exposure and health outcome. The poor quality of epidemiologically-based surveillance systems and disease reporting in many countries sometimes
precludes environmental analysis and hinders the planning and implementationof analytic epidemiologic studies. Analysis of aggregated data at the nationallevel by PAHO has not shown a significant correlation between ENSO and communicablediseases. However, because of limitations in the available data, it ispossible that correlations were present but not detected. Epidemiological studiesand meaningful conclusions on the impact of ENSO or communicable diseaseswill require a long-term temporal and spatial (geographic) perspective with asustained quality of data that quick short term investigations will not provide.International cooperation between scientists, countries, and WHO is required todevelop the global surveillance system required to respond to these global challenges.Although it is impossible to determine how much human illness is directlycaused by ENSO, specific weather patterns give an indication of potentialoutbreaks of vector-borne diseases. It will be very difficult to separate the effectsof El Niño from other factors that impact the transmission of these diseases. Oneprogram that is currently being conducted to address these issues is the ENSOExperiment (Box 2-1).
Given current world population dynamics, including growing populations near coastal waters and increased recreational use of ocean and coastal waters, particularly in developing countries, there are several pressing scientific needs for better understanding the ocean and infectious diseases. The spectrum of infectious diseases in the United States related to several routes of exposure needs to be more thoroughly and comprehensively studied and monitored, nationwide and internationally. These disease categories include: waterborne infectious diseases, vector-borne diseases, infectious diseases related to consumption of tainted seafood, and the infectious diseases related to swimming, boating, fishing, and
water use. New tests are needed to detect waterborne pathogens more quickly and efficiently, and more definitive research in the changing epidemiology of infectious diseases is needed. The primary areas deserving attention are listed below.
1. Newly developed tests, including gene probes for bacteria, viruses, and other pathogens and monoclonal/polyclonal antibody direct detection methods, are available and should be applied to environmental monitoring and quality assessment. However, they must be field tested for application in different geographic locations to expand the database and, if successful, should be adopted by enforcement agencies as standard methods. To achieve this goal, coordination of funding and enforcement agencies, at all levels of government, with industry and academia, will be necessary. Specific suggestions include:
2. Interdisciplinary research should be emphasized/encouraged to assess the potential for global climate change to affect the spread of human diseases. Investigative teams should include, at a minimum, microbiologists, meteorologists, oceanographers, remote sensing experts, epidemiologists, parasitologists, biostatisticians, entomologists, ecologists, botanists, and behavioral scientists. Specific suggestions include: