The Heat Is Online

Health Impacts of the 1997/98 El Nino

The Health and Economic Consequences of the 1997/98 El Nino and La Nina

(Center for Health and The Global Environment: Harvard Medical School)

EXECUTIVE SUMMARY

(The complete report is available at: http://chge2.med.harvard.edu/enso/disease.Html)

Extreme weather events (EWEs) often create conditions conducive to outbreaks of infectious diseases. The upsurge of insect, rodent and waterborne diseases following Hurricane Mitch in Central America in October, 1998 highlights this connection. Heavy rains can produce new breeding sites for insects, drive rodents from burrows and contaminate clean water systems. Conversely, flooding followed by drought can spread fungal spores and spark fires.

The 1997/ 98 El Niño-related extreme weather events spawned "clusters" of disease outbreaks in many regions of the globe. In the Horn of Africa extensive flooding led to large outbreaks of malaria, Rift Valley fever and cholera. In Latin America, extreme weather was associated with outbreaks of malaria, dengue fever and cholera. In Indonesia and surrounding island nations, delayed monsoons and the compounding effects of local farming practices led to prolonged fires, widespread respiratory illness, and significant losses of wildlife.

Throughout the 1990s an area the size of Massachusetts has burned each year in the Amazon. During 1997/ 98 tropical rainforests in Brazil, Mexico and Central America normally soaked and relatively immune to forest fires raged out of control. Winds brought the smoke from tens of thousands of Mexican fires the largest forest fire complex on record for Mexico in a great gyre through the southern U. S.. 500,000 acres later erupted in Florida; 100,000 people were evacuated and 300 homes were lost (Pyne 1998). The losses to Florida in timber alone were over $400 million. Flooding in California spawned agricultural pests and brought large economic losses. In Europe and the U. S., 1998 summer heatwaves killed hundreds.

The 1998 La Niña event beginning abruptly in the Spring of '98 continued the pattern of extreme weather. The impacts have included:

(i) widespread flooding in Bangladesh and China the latter displacing 230 million people;

(ii) flooding in Texas, following drought and a heatwave (with temperatures exceeding 100º F for 60 days 29 in a row and a recordbreaking string of warm nights over 80º F), causing significant agricultural damage as well as human deaths;

(iii) Hurricane Mitch, swamping the Central American isthmus (with thousands of deaths and numerous epidemics);

(iv) a severe October cold wave across Europe; and

(v) a crippling December Ice Storm across the southern U. S.

All told 1998, the warmest year on record and perhaps of the millenium (Warwick 1998) proved to be the most costly year on record in weather-related impacts. In the first 11 months of 1998, $89 billion was lost due to weather related events; more than all the combined losses of the 1980s. If climate change continues to be associated with more frequent and intense El Niño and La Niña events and the accompanying volatile and severe weather patterns we have begun to see the profound consequences climate change can have for public health and for the international economy. There are implications for monitoring, public health early warning systems and for environmental and energy policies.

The El Niño/ Southern Oscillation

The Southern Oscillation is the difference in sea level atmospheric pressure between the island of Tahiti and Darwin, Australia. A negative Southern Oscillation Index (SOI) is associated with slackening of easterly winds (westwardblowing), allowing the western warm pool (WWP) centered east of Indonesia, the world's "hot spot" to shift to the east. A negative SOI is thus associated with an El Niño event, in which anomalously warm water reaches the west coast of South America. This usually occurs around Christmas time, thus the event has been dubbed El Niño, or "The Christ Child".

El Niño events affect the jet stream and are statistically related to extreme weather events in specific areas of the globe (Kiladis and Diaz 1989; Glantz et al. 1991). As examples, Northeast Brazil usually experiences drought, while southeast Brazil and Peru experience heavy rains and often flooding. Southern Africa has a strong "signal," and usually experiences drought.

The opposite pattern is associated with a cold ENSO event (" La Niña"). In a La Niña event often following an El Niño event anomalously cold water appears near South America. Southern Africa usually experiences heavy rains and flooding during La Niña or cold ENSO events. The La Niña event that began in the Spring of 1998 was associated with flooding in Texas and Central America, following extended drought conditions that accompanied the previous El Niño event.

Both anomalous phases with either warm (El Niño) or cold (La Niña) surface waters bring weather extremes to many regions across the globe. With the La Niña phase of 19951996, many regions of the world that had lived with drought during the El Niño years were besieged with intense rains and flooding. Flooding in Colombia, South America, and in southern Africa were accompanied by upsurges of vectorborne diseases. Other areas experienced a switch of the opposite kind, with drought and wildfires replacing floods. A similar pattern of extreme weather events have accompanied the La Niña of 1998, with signals reciprocal to those of the 1997/ 98 El Niño.

Disease Clusters

Disease outbreaks in various regions of the globe have historically been correlated with unusual weather patterns, such as those associated with the El Niño/ Southern Oscillation (ENSO) phenomena (Bouma et al. 1994; Epstein et al. 1995; McMichael et al. 1996; Patz et al. 1996; WHO 1996). In order to explore the associations among disease outbreaks and anomalous climatic conditions during the 1997/ 98 El Niño event, precipitation and temperature anomalies were mapped along with outbreaks of dengue fever, malaria, cholera, Hantavirus pulmonary syndrome and forest fires.

This study is based on correlations rather than timeseries analyses. It is intended as a qualitative search for overlapping of events and as a guideline for a) further modeling and hypotheses testing, and b) pilot programs to evaluate the efficacy of health early warning systems. The sources for this work include primarily ProMED, the World Health Organization's Weekly Epidemiological Record, the Centers for Disease Control and Prevention's Morbidity and Mortality Weekly Review, Emerging Infectious Diseases Journal, and media reports. ProMED the Program for Monitoring. Emerging Diseases is a communication listserv, established by the US Federation of American Scientists, Washington, D. C..

I. BACKGROUND

El Niño events most often begin in the Fall, peak in April or May, and last for approximately one year. The 1997/ 98 El Niño came early, beginning in April of 1997, and ended unusually abruptly in May of 1998. In terms of sea surface temperatures in the Pacific Ocean it was the strongest El Niño event of the century. The El Niño was followed by the sudden appearance of a cold anomaly. This La Niña event, however, was accompanied by persistent warm water conditions in other regions of the Pacific Ocean.

Disease Transmission and Meteorological Conditions

The transmission of malaria and dengue fever (DF) depends upon the life cycle of the mosquito vector and the microorganisms they can carry. Unusually warm weather increases the infectivity of mosquitoes, and increased precipitation provides breeding sites for anopheline mosquitoes that can transmit malaria. (Note: heavy rains can also flush out standing pools and eliminate mosquito larvae.) In mountainous regions, dry periods may result in the pooling of rivers and streams and provide favorable conditions for mosquito breeding. For the periodomestic Aedes aegypti mosquito, which can carry the DF virus, anomalously warm and wet conditions may precipitate outbreaks, while container storage of water can lead to population explosions of mosquitoes during droughts. Where socioeconomic conditions are favorable, mosquito populations and disease transmission may occur only sporadically, under suitable conditions.

In Latin America in 1997/ 98 the outbreaks of malaria and DF occurred in regions that were anomalously warm, suggesting a significant role for warm temperatures in encouraging transmission. Specifically, increases in daily minimum (nighttime) temperatures (TMINs), a hallmark of climate change (see Section IX) (Easterling et al. 1996) that is exaggerated during El Niño events may enhance development of mosquitoes and the parasites or viruses they can carry. Increases in TMINs may also be important in permitting mosquitoes to extend their seasonality and their range to higher altitudes or latitudes, thus exposing new populations with little immunity to increased risk (Epstein et al. 1998). For rodents, droughts can decrease predation while heavy rains provide new sources of food. Flooding can also drive rodents from their burrows, increasing contact with humans, as demonstrated by outbreaks of rodentborne leptospirosis in Central America following Hurricane Mitch.

With cholera, heavy rains can flush organisms (and nutrients and chemicals) into clean water supplies, thus spreading the disease through the water system. In the marine environment, heavy rains may also provide pulses of nutrients that initiate choleraharboring plankton blooms (Colwell 1996; HEED 1998).

Dry extremes are also capable of fostering outbreaks of cholera and other waterborne diseases. Dry streambeds can concentrate microorganisms and impede hygiene. The lack of access to adequate water, in quantity and quality, can also hinder treatment and increase case fatality rates. 8

The World Health Organization Report

The following report appeared in the World Health Organization's Weekly Epidemiological Record on May 15, 1998. It provides an excellent summary of the disease outbreaks associated with the El Niño of 1997/ 98. 9

1997/8 El Niño has already been associated with droughtrelated forest fires originating in Indonesia, which have, in turn, resulted in a dramatic increase in respiratory disease visits in Kuala Lumpur General Hospital and in the State of Sarawak (both in Malaysia). These fires have mainly been caused by human activity but the lack of seasonal rains has led to their spread over wide areas and the fires are now affecting virgin rain forest. Similar fires are being witnessed in the Amazon rain forest and pose a major ecological threat to both farming and traditional indigenous communities. At present droughtrelated famine threatens countries such as Sudan and the Philippines. In 19971998, El Niño has also been associated with very destructive flooding in South America. Ecuador and Peru have been particularly affected.

Infectious diseases: Strong evidence exists of linkages between these weather variations and increases in the incidence of infectious diseases, such as insect vectorborne diseases (e. g. malaria, Rift Valley fever) and epidemic diarrhoeal diseases (e. g., cholera and shigellosis). Climatic factors, such as changes in temperature and humidity, are known to be capable of facilitating or interrupting the capacity of insect vectors to transmit disease to humans. Malaria and Rift Valley fever (RVF) are 2 diseases for which substantial documentation in this area exists. Less well documented, but of increasing interest, are the effects of ENSO on dengue. This largely urban disease, present in tropical regions around the world, is spread by mosquitoes that breed in artificial containers. Thus, in addition to climatic factors, changes in domestic water storage practices, brought about by disruption of regular supplies, will also influence patterns of transmission.

Malaria: El Niño events have an impact on malaria control in many parts of the world because the associated weather disturbances influence vector breeding sites, and hence the transmission potential of the disease. It has been recognised that many areas experience a dramatic increase in the incidence of malaria during extreme weather events correlated to El Niño. Moreover, outbreaks may not only be larger, but more severe, as populations affected may not have high levels of immunity.

Quantitative leaps in malaria incidence coincident with ENSO events have been recorded around the world; such epidemics have been documented in Bolivia, Columbia, Ecuador, Peru and Venezuela in South America, in Rwanda in Africa, and in Pakistan and Sri Lanka in Asia. Historically, in the Punjab region of northeastern Pakistan, the risk of malaria epidemics increases fivefold during the year following a major El Niño, and in Sri Lanka, the risk of a malaria epidemic increases fourfold during an El Niño year. These increased risks are associated with above average levels of precipitation in the Punjab and belowaverage levels of precipitation in Sri Lanka. In South America and Rwanda, heavy rainfall has contributed to major epidemics of malaria. To be able to forecast the impact of El Niño in different endemic areas, control programmes need to develop a thorough understanding of how local vector species respond to climate variability, and how a population's immunity and nutritional status fluctuate over time. To organize a timely and effective epidemic response, malaria control programmes need to incorporate surveillance and epidemic control in their everyday activities.

Rift Valley fever: Outbreaks of Rift Valley fever (RVF), a vectorborne disease that principally infects livestock, have occurred in eastern Africa on almost every occasion that there has been excessive rainfall. As a result of the 1997 El Niño, areas of northeastern Kenya and southern Somalia experienced rainfall which was 60-100 times heavier than normal the heaviest recorded rainfall since 1961. The rains, which began in October 1997 and continued through January 1998, caused RVF virus-infected eggs of floodwater Aedes mosquitoes to hatch. In the outbreak of Rift Valley fever that followed, livestock losses were considerable in the affected regions. Moreover, the estimated toll of human death due to RVF in the region was 200250, while there were an estimated 89 000 human cases of RVF in northeastern Kenya and southern Somalia. Other areas of Kenya, and the United Republic of Tanzania were also affected with widespread animal infections; however, their impact on the human population was not as great. Preliminary estimates of infections and deaths among animals and humans suggest this may be the largest outbreak of Rift Valley fever ever reported.

Cholera and other epidemic diarrhoeal diseases These are a major cause of morbidity and mortality in many countries. Outbreaks can be related either to floods or drought (floods, for example, contaminate water supply, while droughts make hygiene more difficult and contaminate the water that remains). There is circumstantial evidence to indicate a close association between weather changes caused by El Niño and cholera. Since September/ October 1997, there has been a deteriorating cholera situation in the Horn of Africa. After heavy rainfall and floods, most of the countries in this region reported a dramatic upsurge in the numbers of cases of and deaths due to cholera. In 1997, a total of 40 249 cholera cases with 2 231 deaths were reported in Tanzania alone (compared with 1 464 cases and 35 deaths in 1996). Kenya reported 17 200 cases and 555 deaths and Somalia 6 814 cases and 252 deaths due to cholera in 1997. With the floods continuing in this region and adding to already limited sanitation, poor hygiene and unsafe water, conditions favour the spread of cholera.

At the end of 1997 other countries bordering the Horn of Africa, such as the Democratic Republic of the Congo and Mozambique, were reporting increased numbers of cholera cases and deaths. Reported figures through the first 3 months of 1998 already showed 11 335 cases and 525 deaths in Uganda, and 10 108 cases and 507 deaths in Kenya. In the Americas, the current cholera epidemic has been raging for 7 years and, associated with a major El Niño, the number of cholera cases started to increase at the end of 1997.

In 1998, Peru has been suffering from a major outbreak and has already reported, for the first 3 months of 1998, 16 705 cases and 146 deaths. Other countries which are reporting increasing numbers of cholera cases in 1998 are Bolivia, Honduras and Nicaragua. A study examining the relationship between sea surface temperature and cholera case data in Bangladesh during 1994 documented a close association between those two variables.

Prediction and prevention Measures to predict and prevent disease outbreaks related to El Niño are increasing. In southeastern Africa and the Horn of Africa, the regional WHO Cholera Surveillance Teams, warned by early forecasts of El Niñorelated extreme weather events in 1997, were able to help reduce the severity of the cholera outbreak in those regions by means of increased monitoring and heightened preparedness of healthcare institutions in the area. The Southern Oscillation Index has been used to predict the probability of epidemics of vectorborne diseases (such as Murray Valley Encephalitis in Australia). Satellite remote sensing, used to detect areas of abnormal precipitation via increases in vegetation, highlighted exactly those areas which were hit by the RVF outbreak in east Africa in late 1997 and early 1998. The use of mathematic modelling techniques to predict the spread of malaria into new areas in relation to climate changes are also being used (e. g., in Kenya).

More. . .