Introduction:
Climate change presents a significant threat to human health. Increased temperatures can aggravate COPD and asthma and exacerbate the effect of factors such as air pollution or pollen, which are known to affect these diseases. Older people, people with multiple chronic diseases, and deprived populations are affected the most by high temperatures. Quantifying this variation is crucial to inform adaptation to heat policies and to shed light on how effect modifiers such as deprivation, green space, and smoking modify this effect.

Methods:
Daily data on all-cause mortality at the NUTS3 administrative regions (nomos) during 2000-2019 in Greece by age and sex was retrieved from the Hellenic Statistical Authority. The daily mean temperature at 9kmx9km was retrieved from the ERA-5 reanalysis dataset. We employed a case-crossover study design to examine all-cause mortality, cardiovascular and respiratory mortality by age (75<, 75-84, 85+) and sex (male and female), focusing on the effect of heatwaves (deaths occurring during April to September) using six different definitions (durations = >2 or >3 days and thresholds = 90%, 95% and 99% of the annual space specific temperature percentile). We fitted Bayesian conditional Poisson regression models and examined how the effect varies in time and at the NUTS3 region level.

Results:
We retrieved 216,758 cardiorespiratory deaths from April and September during 2000-2019. Most of the deaths were among females (52%). Overall, we observed an increasing effect of heatwaves on cardiovascular and respiratory mortality with increasing duration and temperature thresholds. We also observed an overall decreasing trend of the heatwave effect across the study periods in all heatwave definitions, apart from our most extreme definition, where we observed an increasing trend. We observed weak evidence of spatial vulnerabilities.

Conclusions:
Heatwaves are associated with cardiovascular and respiratory mortality in Greece in the context of a high prevalence of smoking and COPD.