Health and housing consequences of climate-related disasters: a matched case-control study using population-based longitudinal data in Australia

Abstract

Background: Understanding the role of how people are housed in reducing the long-term health and housing effects of climate-related disasters is crucial given our changing climate. We examine long-term health and housing trajectories and health effects of climate-related disasters in relation to housing vulnerabilities over a decade.

Methods: We conducted a matched case-control study using longitudinal population-based data from the Household, Income and Labour Dynamics in Australia survey. We included data from people whose homes had been damaged by climate-related disasters (eg, flood, bushfire, or cyclone) between 2009 and 2019 and matched control cohorts with similar sociodemographic profiles who had not been exposed to disaster-related home damage during this period. We included data from de-identified individuals with at least 1 year of data before disaster and 3 years after disaster. One-to-one nearest neighbour matching was performed on the basis of demographic, socioeconomic, housing, health, neighbourhood, location, and climate characteristics 1 year before disaster. Conditional fixed-effects models for matched case-control groups were used to assess health trajectories, using eight quality-of-life domains on mental, emotional, social, and physical wellbeing, and housing trajectories, using three housing aspects of cost (ie, housing affordability and fuel poverty), security (ie, residential stability and tenure security), and condition (ie, housing quality and suitability).

Findings: Exposure to home damage from climate-related disasters had significant negative effects on people's health and wellbeing at disaster year (difference between exposure and control groups in mental health score was -2·03, 95% CI -3·28 to -0·78; in social functioning score was -3·95, -5·57 to -2·33; and in emotional wellbeing score was -4·62, -7·06 to -2·18), with some effects lasting for 1-2 years after disaster. These effects were more severe for people who had housing affordability stress or were living in poor quality housing before the disaster. People in the exposure group had a slight increase in housing and fuel payment arrears following disasters. Homeowners had increased housing affordability stress (1 year after disaster: 0·29, 95% CI 0·02 to 0·57; 2 years after disaster: 0·25, 0·01 to 0·50), renters had a higher prevalence of acute residential instability (disaster year: 0·27, 0·08 to 0·47), and people who were exposed to disaster-related home damage had a higher prevalence of forced moves than did the control group (disaster year: 0·29, 0·14 to 0·45).

Interpretation: Findings support the need for recovery planning and resilience building to consider housing affordability, tenure security, and housing condition. Interventions might require divergent strategies for populations in different precarious housing circumstances, and policies should target long-term housing support services for highly vulnerable groups.

Funding: The National Health and Medical Research Council Centre of Research Excellence in Healthy Housing, University of Melbourne Affordable Housing Hallmark Research Initiative Seed Funding, Australian Research Council's Centre of Excellence for Children and Families over the Life Course, and Lord Mayor's Charitable Foundation.