October 22, 2018

organic food and cancer

Research published in JAMA Internal Medicine shows that a higher frequency of organic food consumption was associated with a reduced risk of cancer.

A Roundup accompanied this Before the Headlines.

Title, Date of Publication & Journal

‘Association of Frequency of Organic Food Consumption With Cancer Risk – Findings From the NutriNet-Sante Prospective Cohort Study’  JAMA Internal Medicine

Study’s main claims – and are they supported by the data

The study claims that a higher frequency of organic food consumption is associated with a lower risk of cancer. The authors found that the risk of cancer was approximately 25% lower in the quarter of people with the highest organic food consumption compared to the quarter of people with the lowest organic food consumption.

It’s true that the authors have identified an association, but this is only an observational paper (as the authors themselves acknowledge) and this does not demonstrate that organics reduce the risk of cancer.  We can’t be certain that the association is due to organic food consumption – it could be other healthy lifestyle factors that were not accounted for in the analysis, because people who make a point of eating organic food may well take steps to be more healthy in a bunch of other ways.

Although the authors tried to account for a very wide range of variables that might be associated with both organic food consumption and cancer, it is still possible that these could explain the results.

The study does not find a direct link between pesticide consumption in food and cancer risk, and the authors do not make that claim. The authors do hypothesise that the mechanism by which eating organic foods reduces cancer risk is through reduced levels of pesticide residues in organic food-stuffs, but there is no evidence for that provided in this paper.

The study concludes that “promoting organic food consumption in the general population could be a promising preventive strategy against cancer”. This is speculative and not entirely justified because they have not demonstrated a causal link.

The paper is accompanied by a commentary which describes this study very well indeed. The commentary gives a detailed account of the limitations of the study and the conclusions that can be drawn from it.

Strengths/Limitations

Strengths

This is a large and well-conducted study. It is based on a cohort that was set up in part, to study diet and subsequent health outcomes. The cohort in the present study consists of about 69,000 volunteers in France who had access to the internet.  Participants provided information via web-based questionnaires. They were asked to provide access to their medical records for reported health outcomes.

Frequency of organic food intake was assessed consistently for each participant using a web-based questionnaire that asked him or her how frequently they ate 16 different types of organic food.

A very large number of other variables that could possibly account for an association between frequencies of organic food intake were included in the model.

Limitations

This study relies on people self-reporting their own diets.  This can be prone to errors.

The authors do not state whether the questionnaire they used on frequency of organic food consumption had been validated as a reliable method of gathering data prior to the study.

Despite finding an association of higher organic intake with lower risk of cancer overall, the data was only statistically significant for two specific cancers: postmenopausal breast cancer and lymphomas. The data for all other cancers examined by the authors was not statistically significant. The authors only consider a few cancer sites and I don’t think they explain why they have chosen these.

Glossary

Validation: comparison with another dietary assessment method known to give reliable results.

Any specific expertise relevant to studied paper (beyond statistical)?

I have experience of design and analysis of cohort studies, but not of studies of diet or environmental exposures.