Asthma does not respond to the environment in exactly the same way everywhere. A new Australian study suggests that the climatic setting in which children live may substantially alter how vegetation, temperature and air pollution are associated with severe asthma episodes requiring hospital care.
Researchers analysed 11,177 paediatric asthma hospitalisations recorded between 2003 and 2019 across three high-risk local government areas with contrasting climates. Their results show a pattern that is both consistent and complicated: vegetation greenness and temperature repeatedly emerged as important predictors, but the size and direction of environmental associations depended on place and season.
The strongest vegetation association appeared in the tropical monsoon and rainforest region, where higher greenness was associated with a 65% higher hospitalisation rate. In the tropical savannah wet season, sulphur dioxide showed an even larger association, with an incidence rate ratio of 2.11. Yet warmer mean temperatures were associated with lower hospitalisation risk in all three regions.
Why local climate may matter for asthma
Asthma is shaped by multiple interacting exposures. Temperature can influence airway responses, pollen production, mould growth and human behaviour. Rainfall and humidity can change biological allergen levels. Air pollutants can irritate airways and intensify inflammation. Vegetation can provide environmental benefits while also acting as a source or reservoir of allergens.
This makes a single universal relationship between an environmental variable and asthma unlikely. A pollutant concentration recorded during a dry season may occur alongside a very different mixture of heat, humidity, vegetation and airborne biological material than the same concentration during a tropical wet season.
Jialu Wang and colleagues at Queensland University of Technology therefore examined whether environmental associations with paediatric asthma hospitalisation changed across distinct climatic settings rather than assuming that one exposure-response relationship applied everywhere.
Seventeen years of hospital and environmental data
The researchers conducted a retrospective ecological time-series study using statewide inpatient administrative records. They included asthma hospitalisations among children and adolescents from 2003 through 2019 in three local government areas that had previously been identified as high-risk locations and that represented contrasting climatic conditions.
The environmental dataset combined meteorological variables, air pollutants and vegetation greenness. The study drew on established reanalysis and satellite products, including ERA5 and ERA5-Land climate data, atmospheric composition information and MODIS vegetation indices. Vegetation greenness was represented using remotely sensed vegetation information rather than a simple count of parks or trees.
To estimate associations over time, the researchers used distributed lag non-linear models. This matters because an environmental exposure does not necessarily affect asthma hospitalisations on the same day. A lagged model can examine whether an exposure is associated with risk over subsequent days while also allowing relationships to be non-linear.
The team then used random forest modelling as a complementary approach to assess which variables were most informative for predicting hospitalisation patterns. In total, 11,177 hospitalisations entered the analysis.
Greener conditions were associated with higher hospitalisation rates
One of the most consistent findings involved vegetation greenness. Higher greenness was associated with increased paediatric asthma hospitalisation risk across all three study regions.
The strongest estimate occurred in the tropical monsoon and rainforest region. There, higher vegetation greenness was associated with an incidence rate ratio of 1.65, with a 95% confidence interval from 1.29 to 2.11. In practical terms, the model estimated a 65% higher hospitalisation rate at the exposure contrast used by the researchers.
This should not be interpreted as evidence that green space itself causes asthma attacks or that vegetation is generally harmful. Greenness measured from satellites can act as a marker for several processes at once. Lush vegetation can coincide with pollen, fungal spores, humidity and seasonal biological activity. Different plant species also have very different allergenic properties.
The finding is therefore better understood as evidence that broad measures of vegetation can carry locally specific respiratory-health information. A green environment in one climate may represent a different exposure mixture from an equally green environment elsewhere.
Warmer mean temperatures tracked lower risk in all three regions
Temperature showed a more uniform direction. Across all three climatic settings, higher mean temperature was inversely associated with paediatric asthma hospitalisation.
The estimated reduction was approximately 5% to 8% for each 1 degree Celsius increase in mean temperature. That consistency is notable because many other environmental associations changed more sharply across locations and seasons.
Even here, the result does not mean that hotter weather protects children from asthma. Time-series associations can reflect seasonal behaviour, circulating respiratory infections, indoor exposure patterns, allergen cycles and other variables that change alongside temperature. Extreme heat may also carry health risks that are not captured by interpreting the average temperature coefficient as a simple biological effect.
What the analysis does show is that temperature contained substantial information about when hospitalisations occurred. Random forest analyses reinforced this point by repeatedly identifying temperature, together with vegetation greenness, among the leading predictors across climatic settings.
Air pollution effects changed more sharply by place and season
Air pollutants did not produce one stable pattern across the study areas. Their estimated associations were strongly dependent on regional and seasonal context.
The largest reported estimate appeared during the wet season in the tropical savannah region. Higher sulphur dioxide was associated with an incidence rate ratio of 2.11, with a 95% confidence interval from 1.56 to 2.85.
An incidence rate ratio above 2 is substantial, but it should be interpreted in the context of the study design. The analysis is ecological and observational. It identifies a temporal association between environmental conditions and hospitalisation counts, not proof that sulphur dioxide exposure alone doubled an individual child’s probability of being admitted.
The broader lesson may be more important than any single pollutant estimate. If pollutant associations vary with season and climate, asthma surveillance systems based on fixed universal thresholds could miss periods of locally elevated risk.
The same exposure can sit inside a different environmental mixture
The study’s central contribution is its emphasis on context. Environmental exposures rarely occur independently. A child encounters temperature, humidity, vegetation, pollutants and biological allergens together, while those combinations shift over the year.
That helps explain why the same nominal exposure may not carry the same risk signal in every location. Tropical wet-season conditions, for example, can change atmospheric chemistry, vegetation activity and the timing of respiratory triggers at the same time.
For public health, this argues for region-specific forecasting rather than assuming that an exposure-response model developed in one climate can be transferred unchanged to another. Local surveillance could potentially combine weather, vegetation and pollution information to identify periods when children with asthma may face elevated hospitalisation risk.
The results also complicate simplistic interpretations of urban greening. Green space has many established social and environmental benefits, but the respiratory implications can depend on plant species, pollen production, maintenance practices and local climate. Planning for healthier cities may therefore require attention to what is planted and how local allergen seasons behave, not merely how much vegetation is present.
Important limits to what the study can establish
The researchers analysed hospitalisations at the population level, so the study cannot determine each child’s personal exposure. A child admitted with asthma may have spent substantial time away from the residential area represented by environmental measurements, and individual factors such as medication use, asthma severity, housing conditions and socioeconomic circumstances can influence hospitalisation risk.
The ecological time-series design also prevents causal conclusions. Associations with greenness, temperature or pollutants may partly reflect unmeasured variables that move with those exposures. Hospital admissions capture relatively severe episodes and do not represent the much larger number of asthma symptoms managed at home, in primary care or in emergency settings without admission.
Only three high-risk local government areas were examined. Their contrasting climates are useful for studying heterogeneity, but the findings should not automatically be generalised to all of Australia or to other countries. The results are best viewed as evidence that local climatic context deserves explicit consideration in future asthma research and forecasting.
A more local approach to environmental asthma risk
Across 17 years and 11,177 hospitalisations, the study found that environmental associations with paediatric asthma were not geographically interchangeable. Vegetation greenness was positively associated with hospitalisation across the study regions, mean temperature was inversely associated with risk, and pollution estimates changed substantially by place and season.
The findings point toward a more geographically sensitive view of asthma prevention. Instead of asking whether vegetation, heat or pollution is universally good or bad for asthma, the more useful question may be how those exposures combine under a particular local climate at a particular time of year.
Source Information
Study: Wang, J., Cortes-Ramirez, J., Davies, J. et al. “Local climate modifies associations between environmental exposures and paediatric asthma hospitalisation.” Scientific Reports (2026).
Published: 27 September 2026.
DOI: 10.1038/s41598-026-73019-3
Journal: Scientific Reports.
Study design: Retrospective ecological time-series analysis of 11,177 paediatric asthma hospitalisations from 2003 to 2019 across three climatically contrasting Australian local government areas.
