Analyses On Effects Of Meteorological Factors On Incidence Of Scarlet Fever In Beijing And Hongkong

Objective:To analyze the epidemic trend of scarlet fever in China with time series analysis and estimate linear or non-linear effects of different meteorological factors on incidence of scarlet fever,which could provide help for finding suitable condition for transmission of scarlet fever and reference for disease prevention and control of scarlet fever.Materials and methods:Monthly case number of scarlet fever,meteorological factors and number of resident population in Beijing and Hong Kong,from 2004–2015 were collected.Spearman correlation coefficient were used in univariate analysis of relationship between case number of scarlet fever and different meteorological factors.The generalized additive model(GAM)were applied in multivariate analysis for controlling potential effects of time trend and population change.The meteorological effect curve would be estimated for each meteorological factor in two places.Results:There were all 31406 cases were diagnosed with scarlet fever in Beijing,from 2004–2015.The mean number of cases were 218 and mean incidence rate was1.2153/10~5 in Beijing for each month.There were all 7881 cases were diagnosed with scarlet fever in Hong Kong,from 2004–2015.The mean number of cases were 55and mean incidence rate was 0.7791/10~5 in Hong Kong for each month.Generalized linear models with log link function based on negative binomial distribution were applied to data both in Beijing and Hong Kong.After controlling for other variables,there were six meteorological factors showed non-linear effects on scarlet fever in Beijing.The mean temperature range showed negative effect in interval of[5.9,8.5)°C and positive effect in interval of[8.5,13.0]°C.The sunshine duration showed negative effect in interval of[59.3,175.0)hours and positive effect when over than 175 hours.The mean wind speed showed positive effect in interval of[1.6,3.1]m/s,however,the trend of growth became weaker when further growth.The mean relative humidity showed positive effect in interval of[25,45)%and negative effect in interval of[45,77]%.The lowest temperature showed positive effect in interval of[-10,15)°C and negative effect in interval of[15,25]°C.The mean air pressure showed negative effect in interval of[997.2,1015.0)hPa and positive effect in interval of[1015.0,1028.0]hPa.In multivariate GAM model,the lowest temperature,mean air pressure and mean relative humidity were independent risk factors for incidence of scarlet fever in Beijing.For meteorological factors in Hong Kong,all temperature variables showed linear negative effect on incidence of scarlet fever.The RR value was about0.978(0.961,0.995)in when temperature grow 1°C within its range.The wind speed showed positive effect in interval of[3.6,6.5)m/s and negative effect in interval of[6.5,8.8]m/s.The lowest temperature and mean wind speed were independent risk factors for incidence of scarlet fever in multivariate GAM model in Hong Kong.Conclusion:Meteorological factors showed significant effects on incidence of scarlet fever,the difference between Beijing and Hong Kong were partly resulted from their difference intervals on meteorological effect curve.