Association of weather variability with resurging pertussis infections among different age groups : A non-linear approach

Pertussis has resurged in many countries over recent years, especially among adolescents and adults. This study assessed the effect of weather variability on resurging pertussis among different age groups in Jinan, China. Data on weekly pertussis notifications by age group and weather factors (mean temperature (MeanT), mean temperature standard deviation within a week (MeanT SD), diurnal temperature range (DTR) and relative humidity (RH)) were collected between 2013 and 2017. Distributed lag non-linear models (DLNMs) and regression tree models were used to examine the non-linear association between weather variability and pertussis infections. The 2-weeks cumulative relative risk (RR) of pertussis infections was 4.46 (95% confidence interval (CI): 2.33–9.51) in 0–4 age group, 6.25 (95% CI: 1.38–22.76) in 5–9 age group and 10.11 (95% CI: 2.83–39.07) in 10+ age group when MeanT was at 30.0 °C. MeanT SD (RR range in the three age groups: 2.82–5.83), DTR (RR range: 6.33–11.56) and RH (RR range: 2.02–7.43) also exert significant influence, with the highest risks at 10+ age group. Regression tree models showed the interactive effects of weather variability. The mean pertussis infections increased by over 1.7-fold in 0–4 years group when MeanT ≥14 °C, RH ≥57% and DTR ≥10 °C; by over 2.3-fold in 5–9 years group when MeanT ≥20 °C and MeanT SD ≥3 °C; by 2.0-fold in 10+ years group when MeanT ≥0.7 °C, DTR ≥8.3 °C and RH ≥74%. The study found significantly different associations between weather variability and pertussis infections by age group, and appeared to be stronger in 10+ years group. Continuing climate change, together with other risk factors such as low antibody levels among adolescents and adults, may facilitate pertussis resurgence. This supports previous suggestions of carefully reconsidering current vaccination programme to effectively curb the resurgence of pertussis.