Summer school in Treviso, Italy

In June 2023 I got the opportunity to go to Treviso, Italy to attend the Summer School on Modern Methods in Biostatistics and Epidemiology. The summer school aimed to provide introductory and advanced courses in medical statistics and epidemiology and their application to clinical and etiology research and public health. The faculty came from the Harvard T.H. Chan School of Public Health, Karolinska Institutet, Bocconi University, University of Milano-Bicocca, and other European universities. I attended two courses, “Joint Modelling of Longitudinal and Survival Data” by Michael Crowther and Caroline Weibull, and “Causal Inference in Epidemiology” by Michelle Santacatterina.

I wanted to learn how to analyze longitudinal data and use them to identify related health outcomes. Diet is a lifelong habit, and it is useful for all nutritional epidemiologist to gain research skills in longitudinal data analysis. The joint modelling of longitudinal and survival data may help us to understand how dietary components (e.g. lack of vitamin D) over time relate to health outcomes (e.g. development of osteoporosis). Joint modeling in this context would allow us to analyze both the changing/reducing levels of vitamin D intake (longitudinal data) and the time it takes for osteoporosis to occur (survival data). By looking at these aspects together, we can gain a more comprehensive understanding of how diet influences the development of diseases over time. In this course, I learned the methodological framework, underlying assumptions, estimation, model building and predictions.

Dietary recommendations cannot easily be separated from causal conclusions. Causal inferential methods enable us to conclude that a specific dietary component cause or prevent a certain health outcome. The “Causal Inference in Epidemiology” introduced us with framework and tools developed to estimate causal effects from observational data, which most nutritional epidemiologist use. First, the course explained the use of direct acyclic graphs (DAGs), the potential outcome framework, and identification assumptions to identify causal parameters. Then introduced us with statistical methods to estimate these parameters, such as regression adjustment, inverse probability weighting, and doubly robust estimators. All theoretical concepts were set into the context of real-life health research problems. The afternoon lab sessions gave us an opportunity for “hands-on” training in causal inference. By the end of this course, I was able to identify, estimate and compute causal effects using observational data, thus improving my research and decision-making skills.

My PhD project is about the effect of alcohol consumption on cancer risk. During my one week stay in Treviso, I discussed my PhD project with the course leaders and other participants, trying to develop analyzes framework, and this has been a great value for my work. Let’s not talk about the big challenges we discovered for the analyzes of dietary and alcohol data.

Let’s talk about Cison di Valmarino – Treviso. Cison di Valmarino developed from a fortification from the XII century and retains the looks of a medieval town. Cison di Valmarino has a beautiful castle known as Castelbrando where the summer school was held. The castle was surrounded by spectacular green areas, great for hiking during our free time. As a Mediterranean soul living in Norway, I yearned for the taste of good olive oil and ripe, flavorful tomatoes. Every meal was a symphony of taste of the local products produced throughout the province of Treviso. My time there was also a good chance to practice my (almost forgotten) Italian.

Tusen takk to NutriNOR for making this experience possible!