5th March 2014
High protein diets, particularly those high in animal protein, have featured in the media today with headlines suggesting that diets high in meat and cheese may be ‘as bad as smoking’. However, although the headlines are alarming the evidence that they are based on is actually far from conclusive.
These headlines are based on a study published in Cell Metabolism in which the authors suggest that a high protein intake during middle age (50-65 years), particularly a diet high in animal protein, increases risk of cancer and mortality. The paper reports on an observational study in humans and studies of cancer incidence in mice.
Caution should be taken when looking at single, fairly small observational studies in a very complex subject such as cancer and diet, and making direct recommendations from them.
This study used a large cohort of subjects from a prospective study in the US and looked at the association between high protein intake (defined as 20% or more calories from protein) compared to moderate (10-19% calories from protein) and low (less than 10%) protein intakes.
They analysed the data to look for associations between high protein intake and disease/death and, when looking at all subjects over 50 years they found an association with diabetes-related deaths but not with overall mortality. However, when looking only at subjects between the ages of 50-65 years they found a significant association with cancer and all-cause mortality. This association existed for protein from all sources but was higher for animal protein than for protein from plant sources. The authors postulate that this is a causal link and that the reason that protein, particularly animal protein, increases risk of cancer and death is because it increases levels of the growth hormone IGF-1.In contrast there was a decreased risk with high protein diets in the over 65 age group.
It should be noted that such studies can show associations but cannot prove causation as there is no way to be sure that the variable measured (protein in this case) is causing the effect or another factor in the lifestyles of those taking part. This study did control for a number of different factors that could have affected risk of cancer and death but some important factors such as saturated fat intake, physical activity and alcohol consumption were not included. It may be that other lifestyle factors from the small number of people eating a low protein or low animal protein diet may have contributed to the findings - for example, greater consumption of fruit and vegetables.
In this 18 year study diet was measured only once, subjects gave an account of what they had eaten in the last 24 hours at the beginning of the study and this was the basis of their classification into the high, medium or low protein groups and assumed to be representative of their diet for the duration of the study.
Although raised IGF-1 due to increased protein consumption was cited as the likely cause for this association, the study did not find that blood levels of IGF-1 accounted for the association between protein consumption and mortality. They did find that protein consumption significantly increased IGF-1 levels, which is perhaps to be expected as it is known that there is a relationship between IGF-1 and protein intake. However, the effect of IGF-1 on health outcomes is complex and more evidence is needed before the relationship between IGF-1 and cancer is fully understood.
While this study raises some interesting questions, the design does not allow any clear conclusions to be drawn about the effect of protein, or animal protein, intakes in middle age and the mechanism suggested for this link is not well established.
A range of studies in mice were undertaken to investigate the relationship between IGF-1, protein intake and cancer. In both cases mice were fed diets with a high, medium or low protein content and then injected with cancer cells to investigate the progression of tumours. With higher protein intakes, higher IGF-1 levels and greater tumour progression were observed in the mice. However, while interesting, these findings must be interpreted with caution as these experiments do not represent how cancer develops in humans and we cannot assume that people react in the same way as mice to dietary interventions.
It is important to remember that there is strong evidence for being physically active, not smoking, maintaining a healthy weight and keeping alcohol consumption below the daily recommended sensible drinking limits for cancer prevention. This has been supported by high quality systematic reviews of evidence such as the WCRF Diet and Nutrition Report. The relationship between animal protein and cancer risk has not been consistent in all studies, although there is a general recommendation to increase consumption of plant based foods.
This paper does not provide strong evidence for an association between total protein intake, animal protein intake and cancer or mortality. The observational study is not sufficient to demonstrate causation and the suggested mechanism of action is not well established. While there is evidence that high consumption of red and processed meat may increase risk of colorectal cancer, these factors were not investigated or controlled for as part of the study. Similarly, if large quantities of saturated fat from animal products are consumed then this can increase blood cholesterol, increasing risk of cardiovascular disease, but again, this relationship was not investigated in this study.
The animal studies provide an interesting experimental model but the injection of cancer cells into mice does not provide a realistic model of human cancer development.
Average protein consumption in the UK is around 15% of calorie intake, which would have been classified as ‘moderate’ in this study. However, studies of higher protein diets (20-30% calories from protein) have so far not shown adverse effects, although they should not be consumed by those with conditions that affect kidney function. Animal and plant protein sources are included in the ‘meat, fish, eggs, beans and other non-dairy sources of protein’ and the ‘milk and dairy foods’ groups of the government’s eatwell plate healthy eating model and should be included in moderate amounts as part of a healthy, varied diet. Care should be taken to avoid consuming a lot of high fat animal foods as these are high in saturated fat, but animal protein sources can provide important nutrients such as calcium and b vitamins from dairy foods and iron and zinc from meat.