Impacts of Different Grades of Tropical Cyclones on Infectious Diarrhea in Guangdong, 2005-2011

Objective

Guangdong province is one of the most vulnerable provinces to tropical cyclones in China. Most prior studies concentrated on the relationship between tropical cyclones and injuries and mortality. This study aimed to explore the impacts of different grades of tropical cyclones on infectious diarrhea incidence in Guangdong province, from 2005 to 2011.

Methods

Mann-Whitney U test was firstly used to examine if infectious diarrhea were sensitive to tropical cyclone. Then unidirectional 1:1 case-crossover design was performed to quantitatively evaluate the relationship between daily number of infectious diarrhea and tropical cyclone from 2005 to 2011 in Guangdong, China. Principal component analysis (PCA) was applied to eliminate multicollinearity. Multivariate logistic regression model was used to estimate the hazard ratios (HRs) and the 95% confidence intervals (CI).

Results

There were no significant relationships between tropical cyclone and bacillary dysentery, amebic dysentery, typhoid, and paratyphoid cases. Infectious diarrhea other than cholera, dysentery, typhoid and paratyphoid significantly increased after tropical cyclones. The strongest effect were shown on lag 1 day (HRs = 1.95, 95%CI = 1.22, 3.12) and no lagged effect was detected for tropical depression, tropical storm, severe tropical storm and typhoon, with the largest HRs (95%CI) of 2.16 (95%CI = 1.69, 2.76), 2.43 (95%CI = 1.65, 3.58) and 2.21 (95%CI = 1.65, 2.69), respectively. Among children below 5 years old, the impacts of all grades of tropical cyclones were strongest at lag 0 day. And HRs were 2.67 (95%CI = 1.10, 6.48), 2.49 (95%CI = 1.80, 3.44), 4.89 (95%CI = 2.37, 7.37) and 3.18 (95%CI = 2.10, 4.81), respectively.

Conclusion

All grades of tropical cyclones could increase risk of other infectious diarrhea. Severe tropical storm has the strongest influence on other infectious diarrhea. The impacts of tropical cyclones on children under 5 years old were higher than total population.