A number of dynamic models have been produced to explain and predict the daily mechanics of dengue arbovirus transmission to humans by mosquitoes. Such models consider the intricate relationships between numerous meteorological/climatological variables, mosquito biting frequency, breeding vessel water depth, and the dynamics of both human and mosquito populations. As such, prediction of outbreaks of the disease by these models requires a large number and variety of data inputs, many of which are difficult and costly to acquire. The present study uses stepwise multiple regression to relate antecedent thermal and moisture conditions to intraannual and interannual variations in numbers of unconfirmed dengue onset of symptoms cases (1988-1994) for San Juan, Puerto Rico. The goal of this research is to determine the nature of climatic influences on incidence of the disease in this location, and develop a cost-effective predictive tool based on easily obtained climatic indicators for use in public health careplanning.
The mean intraannual trend in dengue onset cases in San Juan, Puerto Rico is represented by a rise in dengue incidence during fall, peaking at the end of the season. A subsequent drop occurs during spring, producing low average incidence in summer. Stepwise multiple regression of mean weekly thermal, moisture, and water budget variables was applied to mean weekly dengue cases lagged for various time increments up to two months. Similarly, these water budget factors were regressed against a trend variable depicting the relative change in dengue cases from the preceding week.
Year-to-year variations in dengue incidence are superimposed upon the mean annual trend, sometimes manifesting themselves as episodes. As in the above analysis, stepwise multiple regression of thermal, moisture, and water budget indicators averaged weekly within each individual year was applied in a stepwise multiple regression analysis against lagged average weekly dengue incidence. This was repeated for standardized values of weekly dengue incidence and a trend variable depicting relative changes in the standardized variable from the preceding week. Results show the relative importance of thermal versus moisture characteristics, and identify climatic characteristics associated with mean intraannual and interannual trends in unconfirmed dengue onset of symptoms cases.