Asymmetric impact of climatic parameters on hemorrhagic fever with renal syndrome in Shandong using a nonlinear autoregressive distributed lag model

Abstract

Abstract Hemorrhagic fever with renal syndrome (HFRS) poses a major threat in Shandong. This study aimed to investigate the long- and short-term asymmetric effects of meteorological factors on HFRS and establish an early forecasting system using autoregressive distributed lag (ARDL) and nonlinear ARDL (NARDL) models. Between 2004 and 2019, HFRS exhibited a declining trend (average annual percentage change = − 9.568%, 95% CI − 16.165 to − 2.451%) with a bimodal seasonality. A long-term asymmetric influence of aggregate precipitation (AP) (Wald long-run asymmetry [WLR] = − 2.697, P = 0.008) and aggregate sunshine hours (ASH) (WLR = 2.561, P = 0.011) on HFRS was observed. Additionally, a short-term asymmetric impact of AP (Wald short-run symmetry [WSR] = − 2.419, P = 0.017), ASH (WSR = 2.075, P = 0.04), mean wind velocity (MWV) (WSR = − 4.594, P < 0.001), and mean relative humidity (MRH) (WSR = − 2.515, P = 0.013) on HFRS was identified. Also, HFRS demonstrated notable variations in response to positive and negative changes in ∆MRH(−), ∆AP(+), ∆MWV(+), and ∆ASH(−) at 0–2 month delays over the short term. In terms of forecasting, the NARDL model demonstrated lower error rates compared to ARDL. Meteorological parameters have substantial long- and short-term asymmetric and/or symmetric impacts on HFRS. Merging NARDL model with meteorological factors can enhance early warning systems and support proactive measures to mitigate the disease's impact.

Keywords

• Hemorrhagic fever with renal syndrome
• Nonlinear autoregressive distributed lag model
• Meteorology
• Asymmetric relationships
• Early forecasting
• Ecological study