Associations between Meteorological Factors and Aseptic Meningitis in Six Metropolitan Provinces of the Republic of Korea

Abstract

We assessed the association between climate factors and a number of aseptic meningitis cases in six metropolitan provinces of the Republic of Korea using a weekly number of cases from January 2002 to December 2012. Generalized linear quasi-Poisson models were applied to estimate the effects of climate factors on the weekly number of aseptic meningitis cases. We used generalized additive and generalized additive mixed models to assess dose-response relationships. A 1 °C increase in mean temperature was associated with an 11.4% (95% confidence interval (CI): 9.6%-13.3%) increase in aseptic meningitis with a 0-week lag; a 10 mm rise in rainfall was associated with an 8.0% (95% CI: 7.2%-8.8%) increase in aseptic meningitis with a 7-week lag; and a 1 mJ/m² increase of solar radiation was associated with a 5.8% (95% CI: 3.0%-8.7%) increase in aseptic meningitis with a 10-week lag. Nino3 showed positive effects in all lags, and its one unit increase was associated with an 18.9% (95% CI: 15.3%-22.6%) increase of aseptic meningitis at lag 9. The variability in the relationship between climate factors and aseptic meningitis could be used to initiate preventive measures for climate determinants of aseptic meningitis.

Keywords: Enterovirus; climate; quasi-Poisson models; sea surface temperature; seasonality.

Figure 1

Map of geographical areas of Korea showing the six metropolitan provinces studied for model. The left panel shows the study sites in Korea map, and the right panel highlights the location of Korea in Asia.

Figure 2

Monthly and yearly distributions of aseptic meningitis cases with meteorological factors in six metropolitan provinces of Korea, time period 2002–2012. The figure indicates monthly variation of aseptic meningitis cases with meteorological factors in the six metropolitan provinces of Korea over an 11-year period. The highest seasonal prevalence is found in the summer months. ^† monthly average of sea surface temperature in Nino3 region; ^§ monthly average of Dipole Mode Index.

Figure 3

Weekly distribution of aseptic meningitis cases with local climate factors in six metropolitan provinces of Korea. The figure indicates weekly variation of AM cases with local climate factors among the six metropolitan provinces of Korea over the 11-year period. The highest weekly prevalence is found in the 28th week.

Figure 4

Quasi-Poisson regression model showing associations between climate variables and the weekly number of aseptic meningitis cases from 2002 to 2012 in six metropolitan provinces of Korea. ^† Weekly average of daily mean temperature (°C); ^‡ weekly average of daily rainfall (mm); ^# weekly average of daily solar radiation (mJ/m²); ^¶ weekly average of sea surface temperature in Nino3 region, ^§ weekly average of Dipole Mode Index.

Figure 5

Generalized additive model for province-specific (A) and generalized additive mixed model for province-combined (B) associations in between the number of aseptic meningitis cases and weekly variations in climate factors in six metropolitan provinces of Korea from 2002 to 2012. * epidemic years (2002, 2005, and 2008); ** non-epidemic years (years excluding the outbreak years). Solids lines in each figure represents log relative risk (RR) at 95% confidence interval (shadows).