PURE and Simple Diet Lowers Mortality, CVD Events

Interviewer: John M. Mandrola, MD; Interviewee: Andrew Mente, PhD


September 13, 2018

John M. Mandrola, MD: Hi, everyone. This is John Mandrola from | Medscape Cardiology, and I'm here at the European Society of Cardiology (ESC) meeting in Munich, Germany. I'm very pleased to have Associate Professor Andrew Mente from McMaster University here, who presented a late-breaking trial today called "Association of Dietary Quality and Risk of Cardiovascular Disease and Mortality in 218,000 People From Over 50 Countries."[1] Welcome Andrew.

Andrew Mente, PhD: Thanks for having me.

Mandrola: Tell us about the study and the topline findings.

Validated High-Quality-Diet Score

Mente: Sure. We found that an all-around high-quality-diet score, comprising fruits, vegetables, nuts, legumes, fish, as well as dairy and unprocessed red meats, was associated with a lower risk for mortality and a lower risk for major cardiovascular events. We applied this in four large international studies and found consistency of the results across studies, and similar results in people with and without vascular disease. We can use this score to make general dietary recommendations for populations globally.

Mandrola: Tell us how you derive this good-diet score.

Mente: We assessed diet using food-frequency questionnaires that were validated in each country, and more specifically, in each urban and rural area of each country in the PURE study. Then we related the dietary information to clinical outcomes events. To develop the diet score, we looked at relationships of foods to mortality. We identified seven foods that were associated with lower risk for mortality in the PURE cohort, and they include those seven foods that I mentioned.

People receive a higher diet score if they consume more of each of these foods. Within the PURE cohort, we found a relationship of [higher diet score] with lower risk for mortality and major cardiovascular events. We applied the same scoring scheme to three other large international studies: a cohort of vascular disease patients (31,000 participants in ONTARGET) and two large international case-control studies of myocardial infarction and stroke (INTERHEART and INTERSTROKE). We found consistency of results across these different studies. This provides further validation of the tool in four large international studies. With this we can make general recommendations on diet globally.

Dietary Quality and Mortality Risk

Mandrola: I noticed from your presentation that the strongest associations with good dietary quality is in overall mortality, and it's less robust in cardiovascular events. Why would that be?

Mente: With major cardiovascular events, we did find a graded association with lower risk for major cardiovascular disease, but it was a more modest association. There was a strong association with noncardiovascular outcomes and total mortality. Noncardiovascular outcomes in PURE largely include cancers and respiratory events. Why is that? Maybe there is some biological explanation behind it, but these are the data that we have and we're reporting it. As we accrue more events in PURE with the ongoing follow-up, we could start looking at individual types of noncardiovascular outcomes.

Mandrola: It would be hardly controversial to say that eating a good diet would improve other kinds of conditions besides heart disease, would it not?

Mente: Not at all. A whole wealth of biological literature [finds] that a healthy diet reduces inflammation and oxidative stress, and provides micronutrients that are very beneficial for cell function and cardiovascular health. It's not surprising at all.

Graded Association

Mandrola: One of the things I'm interested in as a clinician is graded association. In nutrition studies I often see things broken into quintiles and quartiles. But what's the relevance of a graded association in an observational study like this?

Mente: First of all, a graded association more strongly suggests causality. It is one of the Bradford Hill criteria[2] for gauging causality in an observational study. A graded association minimizes the likelihood that the results are due to confounding. Nothing completely rules out confounding, but it certainly makes it less likely that it's due to residual confounding when you see a nice graded association.

Nutrients Versus Foods

Mandrola: The other question that I want to ask you is about dietary quality. You made a comment before about nutrients versus overall food content. Can you comment on this?

Mente: Yes. There has been a growing emergence of people suggesting that we shift focus from nutrients to foods. If you look at trials like the Women's Health Initiative trial,[3] where the focus was on nutrients, the effect was completely neutral. The PREDIMED trial,[4] which focused on certain functional foods, had a much larger effect—about a 30% lower risk for major cardiovascular events and mortality compared with the control diet. We think the focus should be on foods instead of nutrients. When you are at a restaurant, you are not ordering nutrients, you are ordering food. This is easier for the public to translate into diet.

Mandrola: OK, great. Last year, your presentation that saturated fat and overall fat were associated with lower mortality[5] was controversial because it goes against this whole dogma that saturated fats are bad. The PREDIMED trial kind of confirmed that some fat was OK. PREDIMED was retracted but republished and the findings were upheld. But recently an ARIC longitudinal study[6] showed that low-carbohydrate diets were actually bad. Can you comment on whether this is consistent or inconsistent with your findings?

Mente: It's consistent. If you look at the meta-analysis they did in the ARIC study, they identified an optimal level of carbohydrate, which was approximately 55% of total energy intake. In our highest-diet-quality score highest quintile category, the average carbohydrate intake was also 55% of energy. It was an all-around balanced diet, where carbs, fats, and proteins were all in moderation. This is not an extreme diet by any means. This is a diet in moderation and emphasizes several different food groups all in moderation, with a very moderate macronutrient profile.

Diet on a Global Stage

Mandrola: PURE looked at global diets. I understand that there is a difference between a healthy diet globally and a healthy diet in rich Western nations.

Mente: Yes. Certainly, current dietary recommendations were derived from studies conducted a couple of decades back and conducted predominantly in high-income countries. In these countries, the major problem is overnutrition—nutrient abundance. In low- and middle-income countries, there is a problem with nutrient inadequacy—undernutrition. It stands to reason that the findings in high-income countries may not be globally applicable. Therefore, improving the intake of the diet in those populations and increasing the intake of certain nutrient-dense foods may improve their overall nutrition and therefore reduce chronic disease. So, we think our findings are globally applicable. Remember that large parts of the world consume very high amounts of refined carbohydrates. In low- and middle-income countries, for instance, the average carbohydrate consumption was well over 70%, and this is mostly refined carbohydrates. Replacing that with the foods that we've identified—all in moderation—would likely be beneficial to reduce mortality and major cardiovascular events.

Mandrola: Excellent. In the final slide you presented today, you mentioned that there are two aspects of this dietary quality that were controversial. Can you tell us about that?

Mente: There is good consensus that the foods we identified to be beneficial (fruits, vegetables, nuts, legumes, and fish) generally are protective or, at worst, neutral. (With fish it's kind of neutral territory, but fish is thought to be cardioprotective.) So that leaves dairy and unprocessed red meats. We found that up to three servings a day of dairy was optimal, and one-and-a-half servings of unprocessed red meat [daily] is beneficial. This is not an extreme recommendation. It's more of a common sense–type of recommendation. Increasing intake of each in moderation combined with the other foods I mentioned would likely be beneficial.

Mandrola: You are one of the highest-level researchers in the world in nutrition, and the recommendation that I'm hearing is: Moderation, not extremes.

Mente: Absolutely. Eat all of these different foods, all in moderation.

Mandrola: Thank you very much. It's been great talking to you.

Mente: My pleasure.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.