MedWire News: People who eat at least four servings of green salad per week and undertake gardening once or twice a week have a reduced risk for developing lung cancer, regardless of whether or not they are smokers, US study findings indicate.

In a previous study, Michele Forman and colleagues from the MD Anderson Cancer Center in Houston, Texas, examined data from an ongoing case-control study to develop separate epidemiologic risk assessment models for lung cancer in current, former, and never smokers.

Strong risk factors included exposure to second-hand smoke and to dust, a family history of cancer, a history of respiratory disease in the index patient, and smoking history. However, the models only had modest discriminatory power.

In order to include information on diet and physical activity, the team collected data on more than 3800 patients and controls using a 129-item modified National Cancer Institute-Block food frequency questionnaire, with responses categorized in line with the 2006 US Department of Agriculture for food pyramid groups. In addition, physical activity was reported for adult years.

For never smokers, participants who ate three or fewer servings of green salad per week had an odds ratio for lung cancer of 2.09 compared with those who ate at least four servings per week. Similar results were seen for former smokers. For current smokers, the odds ratio for lung cancer was 2.73 for participants who ate three or fewer salad servings per week versus those who at ate least four servings per week.

Gardening also had a significant impact on lung cancer risk. Among never smokers, participants who worked in a garden once or twice a week were 40-46% less likely to develop lung cancer than those who did no gardening. Similar results were seen for former smokers. Among current smokers, the risk for lung cancer was reduced by 33-45% with gardening once or twice a week versus not gardening at all.

The team told the American Association for Cancer Research's Frontiers in Cancer Prevention Research conference in Philadelphia, Pennsylvania, USA, that adding physical activity and diet into the risk prediction model increased its discriminatory power - to 64%, 67%, and 71% for never, former, and current smokers, respectively.

Dr Forman commented: "This finding is exciting because not only is it applicable to everyone, but it also may have a positive impact on the 15% of non-smokers who develop lung cancer."

Meeting website