A diet rich in veg and fish may reduce bowel cancer risk

"Becoming a pescetarian can protect against bowel cancer, new research suggests," the Mail Online reports. The US study found people who mainly eat fish and vegetables, and small quantities of meat, had a significantly reduced risk of bowel cancer.

This study followed more than 70,000 North American Seventh Day Adventists (a branch of Christianity mainly based in the US) for a seven-year period. It looked at whether vegetarian dietary patterns were associated with the risk of developing of bowel cancer.

The study looked at four types of vegetarian dietary patterns:

• vegan – defined as eating eggs, dairy, fish and meat less than once a month (not strictly vegan)
• lacto-ovo vegetarian – more frequent eggs and dairy than above, but still meat less than once a month
• pescovegetarian – eating fish one or more times a month, but all other meats less than once a month
• semi-vegetarian – eating fish and meat one or more times a month, but less than once a week

These definitions are not what most vegetarians and vegans would consider to be truly vegetarian.

Overall, the researchers found people in these vegetarian dietary groups had a combined reduced risk of bowel cancer compared with non-vegetarians (people who eat meat or fish more than once a week).

However, when split into specific vegetarian diet categories, a statistically significant risk reduction for bowel cancer was only found for the pescovegetarian pattern.

Identifying links between specific foods or dietary patterns and consequent outcomes is challenging, as it is difficult to remove the impact of all other health and lifestyle factors. This means that, taken on its own, this study does not prove that fish consumption definitely decreases the risk of bowel cancer.

Still, the results chime with previous studies – there is a broad body of evidence that a diet high in red and processed meat can increase the risk of bowel cancer.

Where did the story come from?

The study was carried out by researchers from Loma Linda University, California, and was funded by the National Cancer Institute and the World Cancer Research Fund.

It was published in the peer-reviewed journal JAMA Internal Medicine.

The Mail Online's reporting of the study was inaccurate for several reasons. The headline "Eating fish but not meat halves the risk of developing bowel cancer" is incorrect. People in the broad pescovegetarian group could also have eaten meat, but not as often as fish.

It is also misleading when the articles state: "Pescetarians, vegetarians and vegans had a lower risk of bowel cancer".

The significant link was only found when the four vegetarian groups were combined, and then only for pescovegetarians when looked at separately. No statistically significant links were found for vegans, lacto-ovo vegetarians, or semi-vegetarians.

What kind of research was this?

This was a prospective cohort study that aimed to look at the link between vegetarian dietary patterns and colorectal (bowel) cancer.

As the researchers say, bowel cancer is one of the leading causes of cancer deaths. Dietary factors are often implicated as a modifiable risk factor.

For example, a review of the evidence (PDF, 556kb) in 2011 by the World Cancer Research Fund (WCRF) concluded there was "convincing" evidence that increased red meat and processed meat consumption is associated with an increased risk of bowel cancer, and increased dietary fibre is associated with a decreased risk.

Vegetarian diets – with their lack of meat consumption, higher fibre content, and the fact adherents often have a lower body mass index (BMI) – might be expected to be associated with lower risk. But the researchers report that this link has not been found for British vegetarian diets.

This large study aimed to investigate different patterns of vegetarian diet and used the most appropriate study design for doing so.

The main limitation with this type of study, however, is that a range of other factors may be influencing any links seen, and removing their effect is difficult.

It is therefore difficult to prove definite cause and effect, although the use of a Seventh Day Adventist cohort should have removed some of these factors.

What did the research involve?

This study was a large prospective cohort of North American Seventh Day Adventists called The Adventist Health Study 2 (AHS-2), which is said to contain a substantial proportion of vegetarians. Almost 100,000 people were recruited between 2002 and 2007.

After excluding people who couldn't be linked with cancer registries, those who reported having had cancer in the past, those aged under 25, or those who had various other missing or improbable data on questionnaires, the researchers had a total of 77,659 people eligible for the study. On average, most participants were in their late 50s.

Dietary information was obtained from a food frequency questionnaire. Using this information, people were assigned into five dietary patterns:

• vegan – consumption of eggs and dairy, fish and all other meats less than once a month
• lacto-ovo vegetarians – consumption of eggs and dairy one or more times a month, but fish and all other meats less than once a month
• pescovegetarians – consumption of fish one or more times a month, but all other meats less than once a month
• semi-vegetarians – consumption of non-fish meats one or more times a month and all meats combined (fish included) one or more times a month, but a maximum of once per week
• non-vegetarians – consumption of non-fish meats one or more times a month and all meats combined (fish included) more than once a week

Cancer outcomes were found through linkage to state cancer registries. They also sent participants two-yearly questionnaires asking about cancer diagnoses.

Various confounding factors taken into account in the analyses included age, gender, ethnicity, BMI, educational level, medical and reproductive history, medication, family history of bowel disease or cancer, smoking, alcohol consumption, and exercise.

In many of their analyses, the researchers combined the four vegetarian groups and compared them with the non-vegetarians. In other analyses, they looked at each vegetarian group separately.

What were the basic results?

Over an average follow-up period of 7.3 years, there were 490 cases of bowel cancer (including cancers of the colon or large bowel and rectum), with an incidence of 86 cases per 100,000 person years of follow-up.

In the fully adjusted model, compared with non-vegetarians, the four vegetarian dietary patterns combined were associated with a reduced risk of bowel cancer (hazard ratio [HR] 0.79, 95% confidence interval [CI] 0.64 to 0.97). 

Looking at the vegetarian dietary patterns separately compared with non-vegetarian diets, only pescovegetarians had a significantly reduced risk of bowel cancer (HR 0.58, 95% CI 0.40 to 0.84). The risk reductions were not significant for the other patterns (vegan, lacto-ovo vegetarians or semi-vegetarians).

How did the researchers interpret the results?

The researchers concluded that: "Vegetarian diets are associated with an overall lower incidence of colorectal cancers.

"Pescovegetarians in particular have a much lower risk compared with non-vegetarians. If such associations are causal, they may be important for primary prevention of colorectal cancers."

Conclusion

This prospective cohort study of a large group of Seventh Day Adventists has examined the links between vegetarian dietary patterns and the development of bowel cancer.

Over seven years of follow-up, it found links between any type of vegetarian pattern overall and a reduced risk of bowel cancer. But when looking at specific sub-groups of vegetarian diet separately, the study only found a statistically significant risk reduction for the pescovegetarian pattern.

This study's strengths are the fact it included a large sample of almost 80,000 adults, and that it linked with cancer registries to look at cancer outcomes, as well as adjusting analyses for a wide range of potential confounders.

However, there are important points to bear in mind:

• Care should be taken before leaping to the conclusion that just eating fish reduces the risk of bowel cancer. The definitions for all four of the vegetarian dietary patterns were quite broad and non-specific. For example, pescovegetarian was defined as consumption of fish one or more times a month, but all other meats less than once a month. This could still encompass a wide range of dietary patterns with variable amounts (and types) of fish, as well as other food groups, such as fruit, vegetables, grains and dairy. It also does not, as the media suggests, exclude people who ate meat – these people just reported eating it less frequently.
• With food frequency questionnaires, it is also possible people provided inaccurate estimations of the consumption of different foods, so they could have been categorised incorrectly.
• Diet was only assessed once at the start of the study, so we don't know whether their diets are representative of lifelong consumption patterns.
• Although the researchers adjusted for many potential confounders, because these were based on assessment at the start of the study only, it is possible the influence of these factors has not been fully accounted for – for example, people's tobacco and alcohol consumption or exercise levels may change. Other unmeasured health or lifestyle factors could also be having an influence.
• The study involved a very specific population group of North American Seventh Day Adventists, who may have distinct health and lifestyle characteristics. This could mean the results do not necessarily apply to other population groups with different characteristics.

This study will contribute to the body of evidence on dietary risk associated with different food types. But on its own it does not prove that fish consumption decreases the risk of bowel cancer.

The World Cancer Research Fund (WCRF), which funded the study, carries out regular reviews of the evidence on risk factors contributing to cancer.

Its last review of bowel cancer was in 2011, and found the evidence on the relationship between fish and bowel cancer risk at that time was limited and inconclusive.

The WCRF will doubtless consider this and any other new studies when it next updates its review, and deliberate on whether this is enough to change the conclusions.

The WCRF currently advises that factors such as the consumption of red and processed meat, alcohol intake, and being overweight or obese are associated with an increased risk of bowel cancer. High dietary fibre, garlic, high-calcium diets and increased physical activity are associated with a decreased risk, they say.

Read more about how you can reduce your risk of bowel cancer.