OBJECTIVE: To analyse the association between types of physical activity (occupational, recreational and household, vigorous and overall) and risk of primary oesophageal (OAC) or gastric adenocarcinoma (GAC). METHODS: From nine European countries, 420,449 participants were recruited between 1991 and 2000 and followed-up for a mean of 8.8 years to register incident GAC and OAC. Information on physical activity (PA), diet, lifestyle and health-related variables was obtained at baseline. Helicobacter pylori infection status was considered in a subset of 1,211 participants. Analyses were repeated by tumour site (cardia/non-cardia) and histological type (intestinal/diffuse). RESULTS: During the follow-up, 410 GAC and 80 OAC occurred. A lower risk of overall and non-cardia GAC was found for increasing levels of a PA index which combined occupational PA with weekly time spent in sports and cycling. The hazard ratio (HR) of GAC was 0.69, 95% CI: 0.50-0.94, for the comparison between active and inactive participants according to the PA index (HR = 0.44, 95% CI:0.26-0.74, for non-cardia GAC). No effect was found for cardia tumours or histological subtypes of GAC. PA of any kind was not associated with OAC. CONCLUSIONS: Overall and distal (non-cardia) gastric tumours were inversely associated with time spent on cycling and sports and a total PA index. No association was found for any type of PA and risk of cardia cancers of the stomach.

Original publication




Journal article


Cancer Causes Control

Publication Date





657 - 669


Adenocarcinoma, Adult, Aged, Esophageal Neoplasms, Europe, Exercise, Female, Health Behavior, Humans, Incidence, Leisure Activities, Male, Middle Aged, Nutrition Surveys, Prospective Studies, Risk Factors, Stomach Neoplasms