Read Original Story

Mediterranean diet and cognitive function: Solid WSJ report would’ve benefited from cost info

Mediterranean Diet Boosts Brain Power, Clinical Study Finds

Our Review Summary

This story does a nice job describing why a study delving into the impact of various versions of the Mediterranean diet on cognitive function is more rigorous than previous efforts. And its provision of strong comments from an independent expert help readers gauge the scope of the potential benefits. There are a few missing components that would have made the story even more useful for readers, including some acknowledgment of the cost of 4 liters of high-quality olive oil per month. We also wonder if there are other ways to potentially slow down the cognitive decline associated with aging that the story could have explored.


Why This Matters

This study offers clinical randomized trial evidence suggesting that a healthy Mediterranean-style diet supplemented with olive oil or nuts may help slow the cognitive decline associated with aging. That’s pretty big news! Diet is implicated in so many health problems, but the relationship is often difficult to prove. Data from randomized controlled trials like this one will go a long way toward solidifying the link.



Does the story adequately discuss the costs of the intervention?

Not Satisfactory

The story mentions neither of two costs that merited discussion: the cost of following a Mediterranean diet and the cost of the study itself. The former is important to anyone who may want to follow the diet if it proves to delay cognitive decline; the latter because it might explain why few such studies exist.

It’s also important to note here that the olive oil and nuts were provided to participants in this study free of charge. And the participants consumed a lot of high-quality olive oil — a liter per week per family. So the costs of replicating this diet would not be trivial, especially for older adults who may be on fixed incomes (i.e. the kind of people enrolled in the study). And that cost might in turn affect how well people outside of the study can adhere to this diet.

Does the story adequately quantify the benefits of the treatment/test/product/procedure?


The story uses numbers (447 cognitively healthy participants divided among three groups) to describe the participants but doesn’t quantify the findings related to the Mediterranean diet with similar precision. Instead it provides this summary: “The results showed that, compared with the control group, memory function remained stronger in the Mediterranean diet plus nuts group, while frontal (attention and executive function) and global cognition benefited in the Mediterranean diet plus olive oil group.”

How much stronger? How did the investigators reach that conclusion?

Numbers are always welcome. And yet in this case, we’re not sure how much it would’ve added for readers to know the change in “cognitive composite Z scores” that were reported in the study. These certainly could’ve been explained and quantified, but we think it was acceptable for the story to turn to a Mayo Clinic neuropsychologist for interpretation instead. That expert tells us, “The changes observed in cognition were very small and didn’t actually show that those diets improved cognition, they just showed less decline.” We’ll call that good enough for a Satisfactory rating.

Does the story adequately explain/quantify the harms of the intervention?

Not Satisfactory

The extra olive oil consumed by participants amounts to more than 500 kcal per day, which could cause weight gain if not balanced by a similar reduction in other food intake. Weight effects weren’t reported here and may not have been a problem in this study. But these were Spaniards already accustomed to consuming a lot of olive oil, vegetables, fruits, and legumes. We don’t know if adding a lot of olive oil to the diets of Americans would yield the same results.

But the story makes no mention of this or other possible harms linked to the diet. If there really are none, that would have been nice to know and would have strengthened the story.

So too would have been a discussion of why, according to the study, more than 25% of the 447 participants “refused to undergo a second procedure or abandoned the study for different reasons.”

Does the story seem to grasp the quality of the evidence?


The story does a good job describing the unique nature of the trial. It makes clear in the second paragraph that the study “employed rigorous scientific practices to test the effect of the diet on health,” and quotes the primary investigator as saying that it is “the first clinical, randomized study using a dietary pattern for good health.”

That’s good enough for Satisfactory rating. But as noted above, the story could have said more about what was measured. Without going into too much detail, something could have been said early on about the fact that the subjects were administered standardized batteries of tests of memory and other brain functions at baseline and after the study.

Does the story commit disease-mongering?


The story does not overplay the import or incidence or cognitive decline.

Does the story use independent sources and identify conflicts of interest?


The story gets valuable input from a clinical neuropsychologist from the Mayo Clinic who was not involved in the research. It also notes the role of the California Walnut Commission in funding some of Dr. Ros’ previous work. We’ll call that good enough for a Satisfactory rating, but we’d note that the donation of nuts and olive oil by commercial concerns in this study is more than a trivial contribution. The acknowledgment of that donation is buried in the original paper under “Additional Contributions,” below mention of the statistics analyst. We think that contribution deserved a more prominent acknowledgment than either the original study or this story gave to it.

Does the story compare the new approach with existing alternatives?

Not Satisfactory

What other dietary patterns might be beneficial for reducing cognitive decline? Are exercise or other lifestyle factors (e.g. education) associated with cognitive benefits as we get older? The story doesn’t explore these questions.

Does the story establish the availability of the treatment/test/product/procedure?


The story describes the Mediterranean diet, which is composed of products readily available in most grocery stores.

Does the story establish the true novelty of the approach?


Very good explanation here. The story notes that “previous observational studies suggested that adhering to a Mediterranean-type diet [is] related to better cognitive function and a reduced risk of dementia.” The story quotes the lead researcher as saying that “this is the first clinical, randomized study using a dietary pattern for good health.”

It’s worth noting that the study authors measured compliance through urine tests of the metabolites of the olive oil and nuts. One reason it’s so difficult to do dietary studies is because compliance is so hard to monitor.

Does the story appear to rely solely or largely on a news release?


Since it includes a comment from an independent expert, we can be certain that this story didn’t rely excessively on any news release.

Total Score: 7 of 10 Satisfactory

Comments (2)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.


May 18, 2015 at 9:20 pm

I have exclusively used olive oil for the last 20years, but certainly not a liter a month! I would quickly regain the weight I’ve kept off for eight years if I did that! I eat nuts as well–in very small quantities. If the trade off for the brain is obesity, I’m going to have to think about it. I would like to think that an overall diet high in plants/fiber and low in salt, fat, and sugar, with the fat being from olive oil and nuts (mostly walnuts and almonds) would be good enough to provide a balance of health vs. disease.



May 22, 2015 at 3:18 am

I’d like to share a few points: one, the conclusion in the abstract states that: “In an older population, a Mediterranean diet supplemented with olive oil or nuts is associated with improved cognitive function.” This statement doesn’t follow from the facts, which really, the peer reviewers should have picked up on. There was no “improved cognitive function”. There simply wasn’t a worsening of cognitive function. Why the peer reviewers at JAMA didn’t pick up on this, I don’t know. Two: there’s lots of stats showing statistical significance, but none that show the clinical significance of this “moderate protective effect”. How moderate was it? Three: where does the “5 tablespoons of olive oil” come from? It’s nowhere to be seen in the study. The study only talks about a diet supplemented with 1 litre of extra virgin olive oil a week. Nowhere in the study does it say that the 1 litre is intended for a whole family. If that’s the meaning of “was supplemented with” can we take it to mean that the 30g of nuts was also intended for the whole family? It’s not right to make claims about the protective effect of 5 tablespoons (75ml) of extra virgin olive oil when it’s not part of the study protocol. Four: if the Mediterranean diet already contains nuts and olive oil, why confuse the story by adding a supplement of nuts and olive oil? As far as I know, there is no specific amount of olive oil or nuts in a Mediterranean diet anyway. To me, this just confuses what could have been a straight forward study comparing the Mediterranean diet to a low-fat diet. (I’d like to have seen an RCT comparing a Mediterranean diet to a diet that wasn’t low in fats, but just contained different fats. Finally, Dr Ros, the author of this study, says in the WSJ piece: “I like walnuts, which are high in omega-3 fatty acids,”. I don’t think the WSJ should have given him this free plug (as your analysis shows, the California Walnut Commission has funded some of Dr. Ros’ previous work. He’s not saying walnuts are better, but given that, in the study, people consumed other nuts too, why choose walnuts as the saviour?