Association between Rainfall and Stroke Admissions: Based on Distributional Lag Nonlinear Modeling

Abstract

Background Stroke is a chronic condition that seriously impairs human health. The correlation between rainfall and onset of stroke remains unclear. Objective To analyze the correlation between rainfall and stroke admissions in Nanchang City, and to provide scientific references for developing a comprehensive prevention and treatment strategy for stroke. Methods Stroke admission data from Nanchang City (2015-2019) from the digital-related group (DRG) system of the Jiangxi Provincial Health Commission Information Center were collected. In addition, atmospheric pollutant data from the national urban air quality real-time release platform and meteorological data from the Nanchang meteorological base station were collected. Basic characteristics of stroke admission patients, air pollutants, and meteorological factors were analyzed. Spearman rank correlation analysis was performed to identify the correlation of case number of stroke admissions with air pollutants and atmospheric factors. Distributional lag nonlinear model was used to explore the linkage between rainfall and stroke admissions. Stratified analysis was conducted based on gender and age (<65 years old and ≥65 years old), and lag represented the lagging days. Results From 2015 to 2019, there were 79 523 hospitalized patients with stroke in Nanchang City, of which 49 072 (61.71%) were males and 48 092 (60.48%) were ≥65 years old, accounting for a large proportion. The number of stroke admissions in winter (December to February) and spring (March to May) were 20 065 (25.23%) and 20 358 (25.60%), respectively. There was a nonlinear relationship between rainfall and stroke admission, and there was a certain lag effect. The RR values of lag1 and lag2 for the effect of rainfall on stroke admission was both 1.009, and 95%CI were 1.000-1.019 and 1.001-1.016, respectively. Stratified analysis showed that the main effect of higher rainfall on the number of male stroke admissions was lag6, RR value was 1.003; the main effect on the number of hospital admissions for female stroke was lag1 and lag2, with RR values of 1.018 (95%CI=1.004-1.031) and 1.020 (95%CI=1.009-1.031), respectively. The main effects on the number of hospitalizations for ischemic stroke under 65 years of age were lag1 (RR=1.016, 95%CI=1.003-1.030), and lag2 (RR=1.018, 95%CI=1.007-1.029) . Conclusion Short-term exposure to higher rainfall can increase the risk of stroke hospitalization, and women and people under 65 years of age are more sensitive to rainfall exposure, and protection should be strengthened for this group of people.

Keywords

• stroke|rainfall|risk factors|distributional lag nonlinear model|hospital admissions|climates and diseases|jiangxi province