Abstract: ObjectiveQuantitative assessment of the relationship between daily average temperature and preterm birth in Guangzhou.MethodsBased on the data of daily preterm births, meteorological and air pollution data in districts of Baiyun and Yuexiu, Guangzhou during 2005-2011, the relationship between daily average temperature and preterm birth was explored by using a distributed lag non-linear time series model.ResultsDuring 2005-2011, there were 263 322 cases of live births including 22 903 cases of preterm births in the two districts.The preterm birth rate was 8.70% (95% CI: 8.59%-8.81%), and daily average preterm births were (9.01±5.61) cases.The daily average temperature was (23.35±6.51)℃, the average daily relative humidity was (66.98±13.98) %, the average daily SO2 concentration was (43.48±28.88)μg/m3, the average daily NO2 concentration was (50.60±50.14) μg/m3, and the average daily PM10 concentration was (80.46±61.61) μg/m3.Associations between daily average temperature and preterm birth were not linear.The exposure-response curve was approximately a U-shape.According to the minimum cumulative relative risk value, the reference temperature was 25.9℃ correspondingly.For the lag period from 0 to 21 days, the cumulative risk effect of daily average temperature was statistically significant beginning at 31℃ (RR: 1.280 3,95% CI: 1.014 9- 1.615 1), and reaching the maximum at 34 ℃ (RR: 1.689 9, 95%CI: 1.076 7 - 2.652 6).There were no statistical significances for the cold effect and cumulative cool effect in the whole lag period.For the thermal effect, when the daily average temperature was 32 ℃ and lag for 1 day, the risk effect was maximum (RR: 1.052 8, 95% CI: 1.014 2-1.092 9); in the lag period from 0 to 14 days, the cumulative thermal effect was maximum (RR: 1.511 7, 95% CI: 1.210 5-1.887 8).ConclusionIn Guangzhou, high temperature weather is an important factor influencing the preterm birth and can increase the risk of preterm birth.