Association of viruses causing acute respiratory infections with meteorological factors and air pollutants in hospitalised children in Macao: a retrospective analysis

Abstract

BackgroundThe association of viruses causing acute respiratory infections with meteorological factors and air pollutants have been reported, but few studies have been done in children in the subtropical marine monsoon climate zone. We investigated the association of such viral infections with particular meteorological factors and air pollutants in infants and children in Macao. MethodsWe searched electronic health records at Kiang Wu Hospital in Macao to retrospectively identify all children aged up to 12 years admitted to the hospital for acute respiratory infection between Jan 1, 2014, and Dec 31, 2017, except for those with congenital respiratory disease, systemic chronic diseases, nosocomial infections, or chronic tuberculosis. All patient data collected were anonymised. Meteorological and pollutant data, including the monthly average of temperature, relative humidity, dew point, and air pollutants (PM10, PM2·5, and NO2) were obtained from the Macao Meteorological and Geophysical Bureau. The primary outcome was the association of the number of respiratory tract infections with these meteorological and pollutant factors. The secondary outcome was determining any lag patterns of such associations. Pearson's correlation and multivariate linear regression were used for data analysis, and multicollinearity between predictor variables was assessed using the variance inflation factor. FindingsOf all 4880 eligible children, 3767 (77·2%) children had tested positive for viruses. 2120 (56·3%) were male and 1647 (43·7%) were female. Their average age was 2·65 years (SD 1·92), with 673 (17·9%) younger than 1 year, 2870 (78·2%) aged from 1 to less than 5 years, and 225 (5·9%) aged 5–12 years. Out of all meteorological factors tested, only relative humidity was weakly associated with the number of viral infections (r=0·143, p=0·038), whereas monthly temperature (r=0·105, p=0·13) and dew point (r=0·131, p=0·059) showed no significant correlation. The overall pollution patterns of the three air pollutants were not similar, but all of them (PM10r=–0·281, p<0·0001; PM2·5r=–0·357, p<0·0001; NO2r=–0·182, p=0·0089) were negatively associated with virus detection. The multivariate linear regression analysis showed that one-week lags in temperature (β=–0·675, p<0·0001) and three-week lags in PM2·5 (β=–0·113, p=0·0059) were also negatively associated with the number of acute viral respiratory infections. InterpretationMeteorological factors could help to explain the prevalence of acute viral respiratory infections. However, air pollutants did not appear to be risk factors for viral infections. Our study warrants further investigation into the influence of the subtropical marine monsoon climate on acute viral respiratory infections in children. An improved understanding of how meteorological changes and exposure to air pollutants could affect respiratory viral infections could help clinicians to decide on diagnoses, prevention strategies, and therapeutic approaches. FundingNone.