By Robert Verkerk PhD

ANH-Intl founder & executive director

Atkins and paleo diets find themselves in the scientific cross-hairs following publication of a new Spanish study published online in Clinical Nutrition.  The study has found that high protein diets increase the risk of death in older adults who are at high cardiovascular risk. It also found that diets in which high protein replaced carbs, resulted in weight gain. By contrast, those where the high protein intake replaced fats, did not. These results, not the first of their kind, have resulted in flurry of negative media headlines bashing high protein/low carb diets. 

About the study

The study conducted by a group of 16 Spanish researchers, headed by Pablo Hernández-Alonso from the Universitat Rovira i Virgili, in Reus, Spain, set out to study the long-term benefits, notably in terms of weight loss, as well as the risks, in terms of all-cause mortality, heart disease and cancer, associated with high protein diets. The study population was a middle-aged and older population at risk of cardiovascular disease (CVD) drawn from a Mediterranean population, specifically the PREDIMED (PREvención con DIeta MEDiterránea) cohort.

PREDIMED is a large, parallel group, multicenter, controlled, randomised clinical trial (RCT) conducted in 7447 older adults (aged between 55 and 80 at the outset) with a high cardiovascular risk. To be included, subjects either had to either have type 2 diabetes or three or more of the following: they were smokers, suffered hypertension, had high LDL cholesterol, were overweight or obese (BMI ≥25 kg/m2), or they had a family history of CVD.  Follow-up was of medium, rather than long duration (2003-2009). Macronutrient intake was assessed using food-frequency questionnaires.

The major finding from the study, that made most of the headlines in the UK, USA and elsewhere, was that high animal protein diets, comprising over 1.2 g/kg body weight (i.e. 84 g protein for a 70 kg adult) increased, rather than decreased, mortality from CVD, cancer and all-causes. 

This equates to protein intakes more than 20% of total energy, as compared with those with an intake of 15-20% of total energy. The study authors also failed to find any benefit of high protein diets in weight loss with high protein intake (from animal and plant sources) being associated with weight gain. Interesting, weight gain was not found when high protein was associated only with animal protein.


The effect of macronutrient composition is complex, and previous trials trying to elucidate health benefits and risks are mixed, often because of big variations in methodology and study population. But of course trials are not the only way in which we can learn about these complex relationships, and many trials ignore key factors that might influence them. Health professionals, lifestyle coaches, nutritionists, sports scientists, celebrities and many thousands of consumers have also been experimenting with higher protein diets. The bottom line is that they have caught on because they work for a lot of people. However, there are a few common adjuncts that are often associated. One is that those engaging with higher protein diets often also increase their activity levels or at least are highly active, and they may also engage in intermittent fasting.

With this in mind, following are some of our key concerns with accepting the bald findings of the Spanish trial:

  1. Missing factors. There are a rash of key factors that influence protein utilisation in the body. These include physical activity levels and patterns, food frequency (e.g. what interval between meals or snacks, was intermittent fasting included?) and effects of other dietary constituents such as phytonutrients. These could have dramatic influences on the way in which proteins are used and metabolised, as well as offering protective effects that have been previously associated with the Mediterranean diet. The older people in the study, being type 2 diabetics or at risk of CVD are unlikely to have active lifestyles. A significant number may also engage in snacking with foods high in refined carbohydrates that will contribute further to their risk of fatal diseases. Dietary polyphenols and other phytochemicals may have a protective effect on cancer and heart disease, and while these have been studied and shown to have an influence in other analyses of the PRDIMED cohort, they were not included in the present study.
  2. Quality and source of proteins and other macronutrients. Animal proteins can involve red meats, white meats, eggs and dairy, while there are a variety of plant proteins, some of which, notably legumes, may be associated with anti-nutritional factors such as lectins. In general terms, there is little or no supporting evidence that white meats or egg contribute to increased rates of disease and the suggestions that high intakes of unprocessed red meat are increasingly being questioned, based on methodological issues with previous studies. However, there are complex issues associated with possible risks and clear benefits with haem (heme) iron in red meat, which may contribute to some degree of risk, while also providing a major source of iron in the diet. An over-supply of sulphur-containing amino acids (methionine, cysteine) may also present a risk. However, based on a balance of evidence, most of the health concerns with animal proteins now appear to be centred on cooking methods or over-consumption of processed and preserved meats (see points 3 and 4 below, respectively). Emerging evidence suggests that this may be associated with specific gene variants/polymorphisms, notably individuals with the rs4143094-TG and -TT genotypes. There is no breakdown of these various types or sources in the study, or consideration of polymorphisms (see point 6, below).?
  3. Cooking/food preparation methods.  High temperature cooking can denature proteins rendering them less beneficial and it can also result in the formation of carcinogens such as heterocyclic amines (HCAs) and polycyclic aromatic hydrocarbons (PAHs). None of these factors have been taken into account.
  4. Additives. Processed meats contain nitrites and nitrates used as preservatives that can form N-nitroso compounds in the colon that are known to be carcinogenic.
  5. Study group may be non-representative of most high-protein diet advocates. The ‘at-risk’ study population was already at some stage of metabolic disease, given all subjects were either type 2 diabetics or had three or more CVD risk factors. Given the relatively short (certainly not long) follow-up, the age of subjects at enrolment, and all of the above factors, including likely low levels of physical activity, it is not possible to suggest any causal effect of high protein on increased risk of CVD or cancer. For example, among the common reasons for people’s compliance with higher protein diets, combined with low intakes of refined carbohydrates, and often higher fat intakes, is their effect on weight loss, especially when combined with regular physical activity. This relationship was not even found in the present study, meaning that its results cannot be applied to such responders.
  6. Genetic variations. The aged, ailing or at-risk study population may have exhibited particular genetic variations (polymorphisms) that exacerbated their risk when protein was substituted for other nutrients, such as fats or nutrient-dense plant food sources, such as coloured fruits or vegetables. Alternatively, they may have polymorphisms that make them more susceptible to certain sources of protein, such as processed meats. Gene-diet interactions are extremely complex and research that has been emerging rapidly with the recent availability of gene sequencing technology is still very limited.
  7. Data collection. Outcomes based on these types of data are of course only as good as the data reported in questionnaires. Self-reporting is notoriously inaccurate, with under-reporting of unhealthy foods, such as high sugar ones or alcohol, being commonplace.

How much protein do you want to consume?

In our view, the present study does little to change our understanding of how diet and lifestyle can be manipulated to optimise health.

General advice, that takes into account the main findings of the Spanish study, includes the following:

  1. Maintain regular physical activity whatever your age
  2. If increasing protein intake from animal sources above 1 g/kg body weight, ensure this does not include significant amounts of processed meat, or meat that has been subjected to high temperature cooking
  3. Substitute refined carbohydrates with multiple sources of vegetables, and ensure these, along with fruit, supply all 5 phytonutrient colour groups daily.

This dietary guidance is fully inline with ANH-Intl’s Food4health plate, which equates to macronutrient ratios from an energy contribution viewpoint of around 25% protein, around 45% fat and 30% carbohydrate.

Find out more in ANH Intl’s four plate shoot-out, which compares two government recommended food plates (UK and USA) against the Harvard Healthy Eating plate and ANH-Intl’s own Food4Health plate.

Related articles