The Spatial Variation of the Effect of Heat on Mortality in Switzerland

I was a co-supervisor of an MSc in Epidemiology student summer dissertation in 2023/24 academic year with Dr. Garyfallos Konstantinoudis as the lead supervisor and Prof. Marta Blangiardo as another co-supervisor. The dissertation was titled The Spatial Variation of the Effect of Heat on Mortality in Switzerland. The dissertation proposal is below.

Dissertation Proposal

Background

Exposure to warm and cold temperatures are established risk factors for human health. A recent study in 854 cities in Europe reported an annual excess of 203,620 (empirical 95% CI 180,882 – 224,613) deaths attributed to cold and 20,173 (17,261–22,934) attributed to heat (Masselot et al 2023). Most of the studies, including the one mentioned above, are focusing on urban environments, neglecting populations living in rural areas. In addition to this, there are few efforts to account for adaptation to temperature. That could stem from pure adaptation to changes in temperature due to climate change, and other attenuation mechanisms due to non-climate factors such as infrastructural changes and improved health care (Vicedo-Cabrera et al 2018).

Aims

The goal of this project is to use available data in Switzerland during 2010-2022 and estimate the number of excess mortalities attributed to warm and cold temperatures accounting for:

a. Differences in age-sex vulnerabilities.

b. Temporal adaptation to temperatures.

Proposed methods

We have already retrieved all-cause daily data during 2000-2023 in Switzerland at small area as retrieved from the Swiss Federal Office of Public Health. We have linked this data with temperature exposure as retrieved from MeteoSwiss. We will use Bayesian spatiotemporal models (as introduced in the Bayesian modelling for spatial and spatio-temporal data module) to quantify the effect of warm and cold temperatures on all-cause mortality. We will account for environmental confounding (e.g., relative humidity) but also use conditional autoregressive priors to adjust for unknown spatial confounding and temporal trends. We will examine the lag effect (an environmental exposure can trigger a health effect also the days following the exposure) of a temperature to all-cause mortality, while using a model that allows a flexible shape on the effect of temperature on health (U-shape).