The world is warming
The 10 warmest years in the historical record have all occurred over the 2014-2023 decade.
When the Heat Strikes
Exploring the impact of extreme heat on older adults in NSW (2009–2019)
Investigator: Zhiyuan Wang
Supervisor: Dr. Heidi Welberry
Background
Australia is no exception
The average number of days above 35°C
increased steadily across 23 NSW meteorology stations during the 2010s,
highlighting a clear warming trend.
Not everyone is equally affected
Older adults are particularly vulnerable due to impaired thermoregulation,
pre-existing health conditions, and social isolation.
Here's What We've Learned So Far
Previous studies have established a clear link between extreme heat and increased morbidity,
emergency department (ED) visits, and unplanned hospitalisations — particularly among
older adults facing vulnerabilities such as
increasing age,
a history of mental illness, or
living in areas of socioeconomic disadvantage.
Identifying the Gaps and Moving Forward
No studies to date have examined the combined effects of age, mental illness,
and disadvantage in an Australian context.
This study explores how vulnerability to extreme heat varies across intersecting risk factors,
including age group, mental health history, and area-level disadvantage.
Methodology
Framing the Question
How does extreme heat exposure impact the likelihood of unplanned hospital admission
and ED visits among older people in New South Wales (NSW), Australia? And how is the effect
modified by history of mental health diagnoses and socioeconomic disadvantages?
How We Studied It
We used a time-stratified case-crossover design, comparing exposure
on the event day with matched control days within the same individual.
This design is ideal for studying transient exposures like extreme heat
and acute outcomes such as emergency visits and unplanned admissions.
By comparing each person to themselves within the same summer period,
we effectively control for seasonality and time-invariant confounders.
Following the Data Trail
Key Variables and Definitions
The Centre for Health Record Linkage links data from the Admitted Patient Data Collection (APDC)
and the Emergency Department Data Collection (EDDC).
The 35°C threshold is defined based on the Bureau of Meteorology's definition
of a hot day.
Area-level socioeconomic disadvantage is measured using the Index of Relative Socio-economic Disadvantage
(IRSD), developed by the Australian Bureau of Statistics (ABS).
Pre-existing mental health conditions were identified based on
the individual's medical records from the past
five years
Bringing Weather to the Map
Interactive Vulnerability Map
Visualizing Vulnerability Across NSW
This map shows areas with overlapping indicators of vulnerability, including mental health ratio, dementia ratio, IRSD percentile rank, and age over 80 rank, based on individuals who experienced unplanned hospitalisation or emergency department visits during the study period.
Descriptive Statistics
This table summarizes key characteristics of a large and representative sample of individuals who experienced unplanned hospitalisation or emergency visits during the study period.
Results
How Heat Affects Hospital and ED Visits
The plot below shows how the risk of hospitalisation or emergency department
visits increases on extremely hot days,
compared to normal conditions.
Extreme heat significantly increases the risk of hospitalisation and emergency visits
The Odds Ratio (OR) compares the likelihood of a health outcome during extreme heat to a normal period.
An OR > 1 indicates increased risk; an OR < 1 indicates reduced risk. Confidence intervals crossing 1 suggest no significant difference.
Extreme heat exposure significantly increases the risk of various health outcomes,
including heat-related hospitalisation (OR = 12.36), all-cause hospitalisation (OR = 1.11),
mental health-related hospitalisation (OR = 1.18), dementia-related hospitalisation (OR = 1.26),
and emergency department visits (OR = 1.08).
Older, disadvantaged adults with mental illness face the highest heat-related risks
We used the ratio of odds ratios (RoR) to assess differences in heat-related health risks across subgroups. Individuals with a prior dementia diagnosis had a 77.4% higher risk of heat-related hospitalisation compared to those without dementia (RoR = 1.774, 95% CI: 0.57–5.57), although this difference was not statistically significant.
For both all-cause hospitalisation and emergency department visits, dementia was associated with a 4% higher risk during extreme heat (RoR = 1.04; 95% CI: 1.015–1.069 for hospitalisation, 1.015–1.066 for ED visits).
Moreover, individuals who were aged 80 or older, lived in the most disadvantaged areas, and had a mental health history faced an 8.7% higher risk of hospitalisation (RoR = 1.087, 95% CI: 1.036–1.140) and a 7.8% higher risk of emergency department visits (RoR = 1.078, 95% CI: 1.031–1.128) compared to their less-disadvantaged counterparts.
Discussion
Who’s Most at Risk?
Mental health history contributed the most to the increased risk.
We observed a clear additive effect among age, socioeconomic disadvantage,
and pre-existing mental health conditions.
What Can Be Done?
An early warning system tailored to vulnerable populations could prevent hospitalisations and save lives.
From Research to Action
Researchers at Griffith University are trialling a system that uses in-home sensors to monitor temperature and humidity.
When levels become dangerous, the system sends alerts to older people along with personalized cooling advice.
Limitations and Future Directions
Although we used a time-stratified case-crossover design to select matched control days,
these days may not always represent valid counterfactuals, as individuals could have experienced other
unmeasured factors that differ from the case day.
While our study provides robust insights, several limitations should be noted. Future research
could benefit from incorporating high-resolution satellite-based temperature data to
improve the accuracy of heat exposure measurement.
END
THANK YOU
Investigator: Zhiyuan Wang
Supervisor: Dr. Heidi Welberry