Background/Objectives/Methods
Epidemiological research to identify lifestyle related cancer causes has been progressively developed during the past few decades, initially based uniquely on case-control studies and, since the past 20-25 years, also benefitting from a growing number of large prospective cohort studies. Several potentially preventive factors (e.g. some dietary supplements) have also been investigated with Randomized Prevention Trials. This large body of research has led to the identification of several factors associated with increased risk of developing specific cancer sites. By far the strongest factors, in terms of relative risk increase are tobacco smoking (with RRs for lung cancer up to 30-fold for highest exposure levels) and alcohol drinking with RRs for head and neck and oesophageal cancers up to 0-fold for high exposure levels. In this context, the identification of the causal role of diet, anthropometry and physical activity has faced the challenge that their association with cancer risk, when found and corroborated by several cohort studies, is quantitatively in a smaller range, typically between 1.5 to 2.5. Therefore, considering how often unhealthy behaviours are correlated, the issue of potential residual confounding from alcohol, tobacco and related lifestyle factors has been one of the key issues in being able to progress our knowledge from “association to causation” for dietary and metabolic factors.  Nevertheless, the consistency in the funding from cohort studies, supported in recent years by finding based on biomarkers and Mendelian Randomization studies, has led to the conclusion that these associations are likely to be causal.  This growing body of  evidence has been analyzed and summarized in the three extensive World Cancer Research Fund (WCRF) reports published in 1997, 2007 and 2018. Over the past three decades, the identification of these risk factors has led several public health bodies at the international and national level to issue public health recommendations, to promote what has, in short, become to be defined as “healthy diet”, as well as promoting increased physical activity and body-weight control.

Results
Primary prevention recommendations, if implemented, are expected to decrease cancer incidence. A major decrease has effectively been observed for lung cancer, firstly in the USA and Australia from the 1980s onward, for both men and women, and more recently in Europe mainly for men, but not yet for women.

Unfortunately, the same has not been observed so far for the cancers more clearly associated with diet, physical activity and overweight/obesity. On the contrary, the incidence of cancers if the colorectum, breast, endometrium, pancreas, prostate and others has been increasing or, at the best plateauing suggesting that these recommendations did not translate in effective cancer prevention.

Conclusions
These observations raise the question of how future public health research should be designed to better understand how recommendations for healthy lifestyle could be effectively implemented in the future to lead to a cancer incidence decrease.