Temporal changes in the high-temperature-mortality association in the Netherlands and the potential impact of the implementation of the national heat plan.

Abstract

Background: To mitigate the health impact of high temperatures, heat plans (HPs) have become widespread around the world. Our aim was to evaluate the temperature-mortality associations and estimate the temperature-related deaths in the Netherlands in the years before (2000-2009) and after (2010-2019) the first activation of the national HP.

Methods: We obtained data about daily all-cause mortality (2000-2019) for the entire Dutch population, and by age, sex, neighborhood socioeconomic status, and urbanization. We linked the daily maximum temperature based on 23 monitoring stations across the Netherlands. Time-series Poisson regression models with a distributed lag nonlinear model, adjusted for long-term and seasonal trends and day of the week, were used to assess relative risks (RRs, 95% confidence intervals [CIs]) in the warm months (May-September). Temperature-attributable mortality fractions for high-temperature exposures and potential HP days were calculated.

Results: We observed positive associations between daily maximum temperature and mortality in 2000-2009 and in 2010-2019. Associations of high temperatures (28.9 °C-95th percentile) were weaker in 2010-2019 (RR: 1.07, 95% CI: 1.05, 1.09) than in 2000-2009 (RR: 1.17, 95% CI: 1.15, 1.20). The attenuation in temperature-mortality risk was strongest for the elderly, women, and individuals living in low-socioeconomic status neighborhoods. The estimated mortality attributable fractions of high temperatures (≥28.9 °C) were lower in 2010-2019 (0.72, 95% CI: 0.60, 0.84) than in 2000-2009 (1.21%, 95% CI: 1.07, 1.33).

Conclusion: The impact of high temperatures on mortality attenuated in the Netherlands. This might be due to the implementation of the national HP, but other factors may have played a role as well.