Does a diet high in protein really cause cancer?
Low fat, high carbohydrate, low carbohydrate, high unsaturated fat, high protein, low glycaemic index……there seems to be never ending advice on what we should and should not eat for optimum health. Often this advice is very contradictory and it is therefore no surprise that the public are generally confused about what they should be eating. The latest addition to this debate is the research on protein intake and cancer which was published in Cell Metabolism and covered by a large section of the media, including this article in the Guardian , under the controversial headline “Diets high in meat, eggs and dairy could be as harmful to health as smoking”.
What does the research say?
This study used data from the National Health and Nutrition Examination Survey, which is a questionnaire based survey of the General Population in the United States. As part of the survey participants completed a 24 hour food recall diary; that is they remembered and recorded what they ate in a 24 hour period. The researchers split the participants into those who ate a high, medium or low amount of protein in their diet and then looked to see whether, in the 18 years after the survey, the participants died and what they died of. They analysed the risk of dying for each level of protein in the diet in two groups – for participants aged 50-65 and those over the age of 65. The study showed that in participants aged 50-65, there was a higher risk of dying (of any cause) in participants who reported eating a higher amount of protein. This risk appeared to be due to animal proteins (meat products) as when they removed the effect of difference in animal protein from the analysis, the higher risk was no longer seen. Yet in participants aged over 65 there was a decrease in risk of death with reported higher protein intake. Similar results were seen for deaths caused by cancer. The study also looked at a proportion of the participant’s blood levels of Insulin-like growth factor 1 (IGF-1). This is a chemical in the body related to the growth of cells. It has been related to the growth of cancer cells as well as healthy cells and studies have shown that higher levels of IGF-1 are linked to higher levels of cancer. Participants who ate higher amounts of protein had higher levels of IGF-1. The researchers think that these higher levels of IGF-1 are the reason why these patients have higher levels of cancer. The conclusion of the study was that having a high level of protein in the diet increases the risk of cancer and death and that it may be due to IGF-1 levels. They suggest that the recommended requirements for protein intake suggested by the Nutritional Board of the Institute of Medicine are decreased, but that older adults may need a higher protein diet.
Are these conclusions fair?
There are a number of flaws in this study which make such strong conclusions and a recommendation to change nutritional guidelines invalid. These types of study – dietary intake and health outcomes are very common and often make the headlines. There are a few simple things to consider when reading the results:
- How good is dietary measurement? Measuring dietary intake is incredibly difficult and full of errors. Participants just do not remember what they ate, with snacking being very hard to recall. The errors are not the same across the dietary spectrum – those with a worse diet (lots of snacking, large portion sizes, processed ready meals) are less likely to accurately record what they ate. It therefore is very unreliable and very prone to bias. This is obvious just by looking at the average calories reported in the study: the average intake was 1823 calories/day, below the recommended amount of 2000 calories/day for a woman with a weight within the healthy range. Given this is from the USA, the most obese country in the world, there appears to be a significant underreporting of intake.
- Confounding factors – is the association causal? Is it actually the high levels of protein in the diet that are causing these results? If you think about what a diet high in protein, particularly protein from meat, looks like you will realise it is not just high in protein but also fat, sugar and salt – think about burgers and pies. It is very difficult to say for certain that the protein is the cause of the increased risk of mortality; it may simply be a marker of a poor quality diet. Also, a poor quality diet is associated with other risk factors such as obesity and low physical activity – these are known as confounding variables. These factors were not accounted for in the analysis.
- Reverse causality- The authors make the assumption that the diet causes the outcomes, but what if the outcome causes the diet? If a participant were ill they are likely to eat less. This may well explain why the association changes dramatically in older adults, who are more likely to have developed chronic illness. If you are unwell you may eat less protein, you are also far more likely to die than someone who is well. That does not mean the low protein diet caused you to die.
What did the media say?
The Guardian reported this study in a very fair and objective manner. It described the study well and had quotes from the senior author restating the conclusion that dietary guidance should be changed. However the article also pointed out flaws in the methodology “easing out the health effects of individual nutrients is notoriously difficult. The apparently harmful effects of a high-protein diet might be down to one or more other substances in meat, or driven by lifestyle factors that are more common in regular red meat eaters versus vegetarians. Other factors can skew results too: a person in the study who got ill might have gone off their food, and seen a proportional rise in the amount of calories they get from protein. In that case, it would be the illness driving the diet, not the other way round”.
The controversy of this research in the media was the headline equating the risk of eating a high protein diet to the risk of smoking. This is not mentioned within the research paper and it is unclear where this came from. It may be that it was issued in the press release from either the journal or the research institution where the research was conducted. This was the focus of much of the negative comment in the Guardian article: “Gunter Kuhnle, a food nutrition scientist at Reading University, said it was wrong “and potentially even dangerous” to compare the effects of smoking with the effect of meat and cheese as the study does. “Sending out [press] statements such as this can damage the effectiveness of important public health messages. They can help to prevent sound health advice from getting through to the general public. The smoker thinks: ‘why bother quitting smoking if my cheese and ham sandwich is just as bad for me?'”
While this study adds to the debate on what is the optimal diet for health, it is not robust enough to influence guidelines. The media appeared to do well in pointing out these weaknesses. It is unclear where the sensational headline came from that equated the risk to smoking, it may have been from the researchers in their press release. This is irresponsible as it misleads the public on important health issues, however it seems to have had the desired effect of gaining the study a lot of media attention.
Levine. M.E., et al. (2014).“Low Protein Intake Is Associated with a Major Reduction in IGF-1, Cancer, and Overall Mortality in the 65 and Younger but Not Older Population”. Cell Metabolism, Volume 19, pages 407-417, doi: 10.1016/j.cmet.2014.02.006.