The PURE Study says it's time to rethink traditional dietary guidelines

Updated: Jan 16, 2019

Dr Salim Yusuf, the President of the World Heart Federation, says high carbohydrates - not saturated fats - are associated with heart disease.

How's that for breaking the glass ceiling of the traditional dietary guidelines?

Dr Yusuf is an accomplished cardiologist and epidemiologist at the forefront of cardiovascular disease (CVD) research and clinical trials. He was a lead investigator in the ground-breaking Prospective Urban and Rural Epidemiological (PURE) Study.

In his presentation of his findings at the Cardiology Update 2017 Symposium, Dr Yusuf speaks out on his recent research findings. He flat out states that the studies that demonise saturated fat are 'fudged', and bluntly criticises the inadequate recommendations for salt intake. It's pretty impressive - here's a prominent doctor, willing to refute the traditional medical paradigm because the research just doesn't back up the mainstream dietary recommendations.

What is the PURE study?

The PURE study is an enormous epidemiological project. In total, it studied 225 000 people for 12 years across 25 countries. Its findings have been published in 60 different articles that focus on a range of topics such as nutrient intake, sodium, physical activity, psychosocial factors of obesity and food labelling practices.

One of the major findings that has attracted the media spotlight is the PURE study on the associations of fats and carbohydrates with total mortality and CVD [A]. CVD is a monumental public health issue. It refers to a collection of diseases of the heart and blood vessels, and commonly includes diseases such as coronary heart disease, heart failure and stroke. About 1 in 5 Australian adults report that they have CVD - that's a staggering 4.2 million people. It accounts for nearly 28% of all deaths in Australia. On average, one Australian dies from CVD every 12 minutes.

The PURE study examined:

  • 135 335 people;

  • Aged 35 to 70 years;

  • From 2003 until 2013, with detailed follow up at 3, 6 and 9 years (median duration of 7.4 years);

  • Across 18 diverse countries on five continents, including low-income, middle-income and high-income countries.

Participants were categorised into quintiles of macronutrient intake based on the percentage of energy supplied from each macronutrient in their diet. The study used country-specific validated food frequency questionnaires and nutrient databases to record dietary intake. The models were adjusted for age and sex, and other factors such as education, smoking, physical activity, waist-to-hip ratio, history of diabetes, urban or rural location and total energy intake.

'Carbohydrates are probably your biggest culprit. So when you eat a hamburger, throw away the bun and eat the meat.'

The PURE population represented a broad range of carbohydrate intake, on average around 46 to 77% of energy. The researchers used the bottom quintile (42 - 49%) as a reference point to assess the impact of a higher carbohydrate diet. The data treats 'carbohydrates' as a broad category and does not capture different forms of carbohydrates.

The findings?

  • A high carbohydrate intake (more than about 60% of energy) is associated with an increase in total mortality and non-CVD mortality.

  • In particular, the data shows a steep increase in the risk of CVD among participants who consumed more than about 55 - 70% of carbohydrate as energy.

  • Higher carbohydrate diets increase forms of dyslipidemia such as elevated triglycerides, decreased HDL, increased small dense LDL and increased blood pressure. This contradicts common attitudes to dietary fats that assume that LDL alone is a predictor of heart disease. [B]

'Contrary to common beliefs, the current recommendations to reduce saturated fats have no scientific basis.'

The PURE population represented a range of fat intake from 8 to 12.6% (quintile 1) to 33.3 to 38.3% (quintile 5).

But fat clogs your arteries, right? Well, not quite. Here's some of the main findings.

  • Dietary fat intake is not associated with higher CVD events. Increasing dietary fats is protective. Higher fat intake is associated with a reduced risk of total mortality, non-CVD disease mortality and stroke.

  • The results actually demonstrated a negative association in quintile 1 (around 8 - 12%) and quintile 2 (around 16 - 19%). That is, too little fat correlates to a higher risk of total mortality, non-CVD mortality and stroke.

The PURE study didn't stop there. It also looked more closely at the different types of fat. But again, the findings don't match up to the current guidelines.

  • Saturated fats (usually cast as the 'bad' fats) are not the villain. They may even be beneficial. And too little may even be harmful. The PURE population studied a broad range of saturated fat intake and looked at people eating a saturated fatty acid intake of < 7% of their diet. The findings failed to identify a significant impact of saturated fat on CVD risk. In fact, the data demonstrates a reduced risk of stroke if carbohydrate is replaced with saturated fat.

  • Saturated fats do not have a harmful effect on blood lipids, and at any rate LDL alone is not a robust predictor of heart disease. This is yet another is a long line of rigorous studies to highlight the fallacies of the 'lipid hypothesis' - that saturated fat and elevated LDL cholesterol are the most important causes of heart disease.[D] The PURE discussion suggests that there are better markers to track CVD risk than LDL.

  • There is 'absolutely no evidence that low fat milk is better for you' than full fat milk. In fact, dairy sources of saturated fat, like milk and cheese, are protective and health-promoting.

  • Monounsaturated fats (ie, olive oil) are clearly protective.

  • Polyunsaturated fats (touted as the 'good' fats, largely from processed vegetable oils) are largely neutral. That said, the PURE study assessed PUFA intake mainly from foods rather than from vegetable oils. This may be a factor that explains the detrimental health impacts attributed to PUFA oils in other studies, but not in PURE.

Based on the PURE findings, the researchers conclude that a total fat consumption of about 35% energy, in combination with a moderate carbohydrate intake, may be beneficial.

In his presentation, Dr Yusuf highlights the dramatic change in the oils consumed in the last 30 years - retailers and dietitians have replaced natural animal fats with vegetable fats that companies can produce and sell. It's confronting to think that profit margins and marketing tactics have dictated our poor dietary choices.

'Stop and think for a minute. Sodium is an essential nutrient.'

In his presentation, Dr Yusuf strongly criticises the current recommendations on sodium intake (2 g to 2.3 g), stating that this is far too low.[C] Dr Yusuf explains that sodium is an essential nutrient and plays many vital roles in the body that are not commonly talked about. For example, sodium is the first line of defense against infection in the skin, and that's why the largest part of sodium in the body is stored in subcutaneous fat. Dr Yusuf gives an example of this - if you take an animal model and make a wound in one limb but not the other, within one hour the level of sodium at the site of injury doubles to prevent infection.

Dr Yusuf notes that the medical community have used the data from the Yanomami Indians to show that lower sodium intake means lower blood pressure. Sounds plausible. But as Dr Yusuf points out, the Yanomami have a life expectancy of 32 years. And die off not from heart disease but from infection.

The findings?

  • A moderate amount of 3 to 6 g daily is optimal.

  • Any less than 3 g a day 'creates uncertainty'.

  • More than 5 g of sodium may be harmful but only in people who have hypertension and are 'sodium sensitive'.

Why should I care?

Epidemiological studies do not establish causation, but large data sets can be a legitimate basis for claims that one thing does not cause another thing because there is no evidence of a correlation. In the PURE study, the findings build a strong case for the proposition that saturated fat does not cause heart disease.

Even if you're sceptical, this study should at least cause you to stop and consider the elephant in the room - why is it that time and time again we see studies and findings that completely and utterly contradict our current dietary guidelines and the mainstream nutrition mantras? And this study is not alone - the PURE data adds to a large and increasing body of evidence that increasing dietary fats are not associated with higher CVD or negative health outcomes. Quite the opposite - we tend to see a trend that saturated fats are vital to our diets, and in amounts far higher than the suggested 10% cap.

It is also a potent reminder to look at the bigger picture. While the medical lens routinely applies a limited approach and tends to blame saturated fat in the diet for elevated cholesterol and possible heart disease, this lipid hypothesis is tenuous and discredited. The findings in PURE highlight that it is futile to try to predict the clinical effect of a single nutrient like saturated fat on CVD. If it sounds too simple, it usually is.

Does PURE support a LCHF diet? No. It's not meant to. That's not the purpose or the scope of the PURE study. The PURE findings focus on the consequences of high amounts of carbohydrate in the diet. But putting aside your ideas about optimal macro ratios, the PURE findings matter because mainstream dietary guidelines tell us that we should eat plenty of carbohydrate-dense foods. The guidelines suggest that a diet that contains 60% or more of energy in the form of carbohydrate is perfectly acceptable, if not ideal. The PURE findings stand at odds with this dietary dogma.

Dietary guidelines need to be reconsidered. It's time to remove the current restrictions on fat intake and instead confront the detrimental effects of high carbohydrate intake.

What do the Australian dietary guidelines say?

The current Australian Dietary Guidelines outline the following 'Acceptable Macronutrient Distribution Ranges' related to reduced risk of chronic disease':

  • 20 - 35% of total energy intake from fat

  • 45 - 65% from carbohydrate

  • 15 - 25% from protein.

How about saturated fat? The message is clear - limit it and choose reduced fat options. Saturated fat is demonised and cast aside along with trans fats. The Guidelines recommend that:

  • Total saturated fatty acids and trans fats comprise no more than 10% of energy intake.

  • It also states that 'continuing research into diet and cardiovascular disease emphasises reducing saturated fat in the diet, which means limiting intake of foods with high saturated fat content.'

This is standard. But the PURE findings clearly stand at odds with a recommendation to limit saturated fat to less than 10% of intake. And its not alone.

Maybe it's time to clear the unfairly demonised saturated fat of its charges.


[A] M Dehghan et al, 'Associations of fat and carbohydrate intake with cardiovascular disease and mortality in 18 countries from five continents (PURE): a prospective cohort study.' The Lancet 2017 390 (10107); 2050-2062.

[B] A Mente et al, 'Association of dietary nutrients with blood lipids and blood pressure in 18 countries: a cross-sectional analysis from the PURE study.' The Lancet Diabetes & Endocrinology 2017 5 (10): 774-787.

[C] See also M O'Donnell et al, 'Urinary sodium and potassium excretion, mortality, and cardiovascular events.' N Engl J Med 2014 Aug 14; 371 (7): 612-23.

[D] For a useful commentary on the lipid hypothesis and a recap of the science behind it, see Sally Fallon Morell's book 'Nourishing Fats', Michael Pollan's book 'In Defense of Food' and Professor Tim Noake's essay, 'Scientific Justification for a Banting Diet.'