Identifying impacts of extreme weather events on mental health in the Republic of Ireland using the Impact of Event Scale-Revised (IES-R) index and machine learning

Abstract

Extreme weather events (EWEs) have become a significant concern due to the global effects of climate change, particularly regarding their impact on mental health and associated direct and indirect healthcare costs. This study explores the mental health impacts of EWEs in the Republic of Ireland, using the Impact of Event Scale-Revised (IES-R) to assess trauma and stress. A cross-sectional survey was conducted across Ireland employing two-step cluster analysis, generalized linear modelling, and regression trees (rpart) to identify psychological stress ‘clusters’ based on verified mental health and well-being measures. Four psychological stress clusters (‘high 33.8 % n = 154’, ‘moderate 21.2 % n = 96’, ‘mild 18.9 % n = 86’, and ‘low psychological stress 26.3 % n = 120’) were statistically identified with the ‘high psychological stress’ cluster having the highest summed IES-R score (59) and the ‘low psychological stress’ cluster having the lowest (5). Members to the ‘high psychological stress’ were less likely to have suburban residence (OR = 0.31), graduate (OR = 0.32) and postgraduate (OR = 0.37) educational attainment, and more likely to have reported poorer health (OR = 1.91) and worsened financial situation (OR = 1.95) post-EWE. Conversely, ‘low psychological stress’ cluster members were less likely to have experienced personal injuries (OR = 0.29) or a worsened financial situation (OR = 0.28) post-EWE and were more likely to be older (>65 years of age) (OR = 5.42), retired (OR = 6.21), have a post-graduate educational level (OR = 4.19), and suburban residence (OR = 3.75). Machine learning models demonstrated a relatively accurate fit for predicting ‘low psychological stress’ membership (AUC = 0.74), with EWE-related injuries, age, EWE type/recency, and occupation as primary predictors for cluster membership. Results show that temperate climates like Ireland may experience milder physical impacts of climate change compared to other regions. The study addresses an important research gap by employing innovative machine-learning techniques to identify patterns in climate-related mental health issues. The findings can help inform evidence-based decision-making, allowing for targeted interventions—both public and private—to improve mental health outcomes for vulnerable populations affected by EWEs in the ROI and similar regions.