CLIMATE VARIABILITY AND PATTERNS OF DENGUE IN THE CARIBBEAN

 by Dharmaratne Amarakoon, Anthony Chen & Michael Taylor,
Climate Studies Group Mona, UWI,
& Samuel Rawlins, Caribbean Epidemiology Centre, Trinidad & Tobago

Introduction

It is a known fact that millions of people are under the threat of dengue fever in the tropics and subtropics. Even though vector-borne diseases such as malaria have been eradicated from the Caribbean, over the period 1980-2001 43,000 cases of dengue were reported. About 41,000 cases have been reported in the last decade, and indications are that the frequency and the intensity of outbreaks of the disease are on the rise.

Studies done in other parts of the world have inferred an association of upsurges of dengue with ENSO events, through temperature increases and availability of stored water during droughts, as well as stagnant water after rains. These conditions enhance mosquito breeding and dengue transmission rate.

Recently published studies (Chen & Taylor, 2002; Peterson et al., 2002) indicate drier than normal conditions with warmer temperatures in the latter half of El Niño years, an increase in precipitation in the early part of the following year, as well as a warming trend in the Caribbean. Thus it appears that we can expect a higher probability of dengue outbreaks in the Caribbean in El Niño or El Niño + 1 years, as more favourable conditions exist for vector breeding as well as disease transmission in such years.

The work summarized in this paper presents the pattern of dengue, its seasonality, and the level of association of the epidemic with ENSO events in the Caribbean. The work encompasses the preliminary findings in the retrospective component of the AIACC project SIS06: The Threat of Dengue Fever – Assessment of Impacts and Adaptation to Climate Change in the Caribbean.

Data and Methods

 The study period was 1980 to 2001. The climate data analyzed were daily and monthly values of maximum and minimum temperatures, and precipitation. The climate data were available in the Climate Studies Group Mona depository. The Caribbean Epidemiology Centre provided dengue and vector data.

Data analysis consisted of: time series analysis of annual reported dengue cases and their rates of change; time series analysis of temperature, precipitation, and anomalies of temperature and precipitation; study of climatology of temperature and precipitation, and monthly average variation of reported cases; and performance of statistical significance tests for observed correlations, where applicable.

ENSO year classification was according to the NOAA-CDC MEI index. El Niño years considered were: 1982-83, 1986/87, 1992/93, 1994/95 (weak), and 1997/98. La Niña years considered were: 1988/89 and 1998+/2000.

Results

 Figures 1, 2 and 3 and the table present the results. In Figure 1, the rate of change represents the change in the reported cases over one year. It is assumed that the population has not increased significantly over the two decades. Figures 2 and 3 are provided as sample cases, and they are for Trinidad and Tobago. Trinidad and Tobago contributed 50 per cent of the reported cases, and the figures illustrate the general pattern of variation of temperature and precipitation. Note that values with a single * in the table correspond to the situation where 1994/95 was treated as a weak El Niño event. The double ** corresponds to the situation of > 90 per cent statistical significance, without the weak 1994/95 El Niño event.

Discussion

 In general, across the Caribbean region the last decade is observed to have been more prone to the disease than the previous decade. There is a periodicity of about four to three years in the previous decade, compared to three to two years in the last decade. The intensity and frequency of the outbreaks in the last decade are more than those in the previous decade.

These features may be due to the fact that, in the last decade, temperatures were higher and rainfall was less abundant, as indicated by the behaviour of temperature and precipitation anomalies for Trinidad and Tobago. These are favourable conditions for reducing the vector incubation period and increasing the disease transmission rate.

Of the eight outbreaks observed, three occurred in El Niño years and three in El Niño + 1 years, including 1995. Thus the probability of an epidemic in an El Niño or an El Niño + 1 year is high compared to that in a La Niña or a neutral year. As seen in the table, the association of the epidemic in the region with El Niño and El Niño + 1 years is statistically significant at the 94 per cent level.

Further, the epidemic shows a well-defined seasonality, as seen in Figure 1a. It occurs in the latter half of the year, during which the temperatures are higher and the rainfall less abundant, especially under El Niño conditions.

It is worth mentioning that, even though details are not presented in this report, the onset of the epidemic in an El Niño + 1 year appears to occur earlier, by about two months, than the occurrence in an El Niño year. In an El Niño + 1 year, higher early season rainfall, which leaves stagnant water in discarded objects, may provide suitable habitats for enhanced mosquito breeding, hence triggering the epidemic with the onset of the warm Caribbean summer.

Summary

 It is clear from the results that an association exists in the Caribbean between dengue epidemics and ENSO events. The seasonality seen here, and the predictability of ENSO events, could provide useful tools in designing early warning systems. Needless to say, socio-economic factors may also influence the triggering of the epidemic.

References

Chen, A.A. & M.A. Taylor (2002). ‘Investigating the link between early season Caribbean rainfall and the El Niño + 1 year.’ Int. J. Climatol., 22, 87-106.

Peterson, T.C. et al. (2002). ‘Recent changes in climate extremes in the Caribbean region.’ J. Geophys. Res., 107(D21), 4601,doi:10.1029/ 2002JD002251.