Walnuts, Cognition, and Modern Science

Here's a party trick. Challenge someone to name a medical condition that hasn't been treated with the walnut plant. Ask them to choose a condition laypeople would know. Then look it up.

Chances are, the bark, leaves, shell, or seed itself have been used by someone, somewhere to treat the condition.

Cardiovascular disease, cancer, diabetes, immune disorders, digestive issues, skin problems, tooth decay, impotence...the list seems endless.

Folks also rely on the walnut to enhance well-being. In some of the most remote areas of the Kashmir Valley, the locals boil walnuts in milk, add a little honey or sugar, and eat the concoction early in the morning as a "brain tonic."

This newsletter focuses on whether the Kashmiris are on to something. I'll be discussing a new study suggesting that eating walnuts with breakfast improves cognitive functioning throughout the day.

I was excited to discover this study. I often eat walnuts in the morning, because I've noticed that doing so helps keep me alert. But I'm a sample of one, and I may be wrong.

Meanwhile, the new study is informative, but I'm not sure it justifies treating walnuts as "brain food".

Folk medicine, modern science

From a modern scientific perspective, folk medicine seems more accurate about the hows than the whys.

In other words, lay people may know how to enhance their health and well-being without understanding the biochemical basis of the enhancements.

A famous example is the Native American societies who chewed willow bark, or drank tea brewed from it, to alleviate pain. Modern science tells us that our bodies convert salicin in the bark to salicylic acid, the active ingredient in aspirin.

As for "brain foods", one of the more common ones is the coca leaf, chewed by millions of people in the Andes to ward off fatigue. The practice is at least 8,000 years old, but ancient Peruvians couldn't have known that chewing the leaf releases low, slowly absorbed levels of alkaloids such as cocaine.

We may know about salicylic acid and tropane alkaloids and such, but mainstream science doesn't necessarily grasp the hows, much less the whys.

For instance, according to the medieval doctrine of signatures, plants contain observable features, or "signatures", that resemble the body parts they can treat. Because walnuts look like brains, medieval physicians prescribed them for whatever they took to be brain-based conditions.

(Some folks see a resemblance between brains and the nut itself, but the shell was often viewed as more similar. If anything, the nut is reminiscent of the ravages of severe Alzheimer's disease.)

The doctrine of signatures lives on in homeopathic medicine, but it's not supported by scientific evidence. Sliced mushrooms may look like ears, but they're not going to help with my hearing problems. Walnuts may be good brain food, but not due to any physical resemblance.

We tend to assume that modern science surpasses both folk medicine and the science of bygone eras at teasing out the hows and whys. Thanks to statistically-based methods, we know how to improve our health and why our "tonics" work.

I want to discuss the new study and then revisit this assumption. My conclusions may surprise you.

Walnuts with breakfast

The new study, led by Lynne Bell at University of Reading (UK), appears in the upcoming issue of Food & Function and is available now online.

Before getting to the details, I should mention that the study was funded by the California Walnut Commission.

I wouldn't assume any conflict of interest, given the pervasiveness of industry-funded research, but it's worth pondering the question. I reached out to Lynne Bell, the lead author, for her perspective. Here's what she shared:

"Much of the current research on nutrition-related benefits to cognition relies on funding from the food industry. These types of study are expensive to run, and other funding opportunities..., are highly competitive....

In many cases, as here, the researchers will have a particular research interest which they would like to pursue. They then design a study and submit a proposal to the industry funder...The funders choose whether or not they would like to fund the research, but don’t play a role in conducting the research [as in the case of this study]..."

In short, the researchers designed and ran the study themselves. The study itself gives me no reason to question their impartiality.

Study methods

In some ways this is the best kind of study – rigorous control, realistic conditions. Plus free food!

Here's an overview:

• Participants came to the researchers' lab one morning before eating anything.

• Blood samples and EEG recordings were taken. Participants took tests of memory and executive functioning (i.e., skills that help us plan and achieve goals).

• Half the participants were given breakfast containing walnuts; the other half received a walnut-free breakfast.

• Participants remained in the lab for six hours. EEGs and tests of cognitive functioning were re-administered 2 hours, 4 hours, and then 6 hours after breakfast.

Some studies lose credibility when you dig into the methodology. This one just gets better. Here are three examples:

—Participants were fairly homogeneous (18 to 30 years old, BMI < 30, nonsmokers, etc). In the day or days prior to the experiment, they agreed not to consume alcohol, caffeine, or foods high in omega-3 fatty acids and/or flavonoids (nuts, oily fish, etc.).

—The breakfasts served in the lab were almost identical – vanilla yogurt and muesli – except that the walnut group received 50 grams of walnuts (about a handful), while the control group's yogurt contained a small amount of melted butter to match the fat content of the walnuts. All participants were served the same lunch.

—The participants came to the lab a week later and repeated the experiment. If they'd been in the walnut group previously, they now participated in the control condition, and vice versa.

The sample was small – 32 participants – but this is a minor limitation given the extent of experimental control and multiple testings of participants. A technical procedure called a power analysis indicated that the sample size was suitable for a conventional statistical approach.

Some key findings

• At the 6-hour mark, the walnut group showed better recall memory than the control group did.

• The walnut group showed faster reaction times on executive function tasks. The difference was observed after breakfast throughout the day.

In each case the differences were small but significantly different. Hooray for walnuts!

This is where the story ends for the national news media. For instance, here's Newsweek and The New York Post:

"The research revealed that eating 50 grams of walnuts...mixed into muesli and yogurt led to faster reaction times and better memory performance later in the day...

"The results were clear. Those who ate the walnut breakfast in the morning had quicker reaction times and better memory performance later in the day compared to the nut-free breakfast."

I would love to end this newsletter here, being a fan of walnuts. But the results simply aren't "clear".

Here are what I consider the two most serious concerns.

The multiple comparison problem

As I mentioned, a few of the researchers' analyses yielded small but significant differences.

6 hours after breakfast, for instance, the walnut group recalled 10.75 words on average, while the control group recalled 10.32. (This is after hearing a list of 15 words read out loud.)

However, most of the more than two dozen memory-related analyses showed no group differences.

The concern here is that when so many analyses are run, a few are bound to turn up significant, just by chance.

Imagine tossing a coin 5 times and recording the results. Then you repeat the process. Keep on doing this and you'll get 5 heads in a row on rare occasions. But that's not grounds for supposing that the coin is weighted, or otherwise special.

(The researchers used something called a Bonferroni correction to address the multiple comparison problem, but such corrections only reduce the chances of the problem occurring rather than actually preventing it.)

Is this a problem only with the memory tests? No. In her email to me, Dr. Bell graciously confirmed that even though the walnut group had faster reaction times overall, the difference was not observed for every test at every testing (i.e., 2, 4, and 6 hours after breakfast).

The mechanisms problem

Assuming that walnuts improve cognitive functioning, why do they help?

Blood serum analyses revealed no group differences, but the EEG data did. For instance, the walnut group showed more alpha, delta, and theta wave band activity in the frontal lobe, where neural resources would be used for the memory tasks.

Unfortunately, the EEG data isn't very meaningful, given that the walnut group didn't consistently outperform the controls.

That is, if walnuts improve brain functioning in areas responsible for memory, it's unclear why the walnut group would fail to outperform the controls on most of the memory tasks (and actually perform more poorly on one at the 2-hour mark).

Imagine that two chess masters play each other 100 times. Master A wins 3 matches, Master B wins 1 match, and the remaining 96 matches are draws.

Imagine too that you've identified an important difference in pre-match preparation. Master A always exercises. Master B always naps.

Should we conclude that exercise offers better preparation for a chess match?

No. Because we wouldn't have a good explanation for why draws occurred 96% of the time.

Likewise, the EEG differences don't seem to account for the cognitive performance of the walnut group, because most of the time their performance wasn't affected.

What does the data tell us?

The study's senior author, Claire Williams, released a statement on the University of Reading website that was quoted by Newsweek, The New York Post, and others. The statement is, indeed, very quotable:

"This study helps strengthen the case for walnuts as brain food. A handful of walnuts with breakfast could give young adults a mental edge when they need to perform at the top of their game..."

Here's my less catchy takeaway from the study:

1. It's unclear whether walnuts enhance cognition.

In this newsletter I've discussed a few of the reasons for skepticism about the findings. I hope I'm wrong. The study is brilliantly designed, but the data just doesn't support that "mental edge" claim.

Other studies suggest that walnuts benefit cognitive functioning, but I agree with Bell and colleagues that their results aren't conclusive either. So, the best-case scenario is a tentative "maybe".

2. It's unclear why walnuts would enhance cognition.

Let's say you choose to trust the data. What is it in walnuts that improves recall memory and reaction time?

Some research implicates omega-3 fatty acids and/or flavonoids.

In Bell and colleagues' study, protein is another possibility. Although the walnut and control groups were served nearly identical breakfasts, the walnut breakfast contained about 5 more grams of protein. (That seems like a meaningful amount – about 10% of the daily minimum recommended for a 140 pound person.)

3. It's unclear whether anyone needs walnuts.

Our brains (and bodies) can't function without omega-3 fatty acids and protein, but we can obtain them from many foods. Same goes for flavonoids, which are not essential nutrients.

Should we eat walnuts with breakfast?

As I mentioned, modern science is often considered superior to both pre-modern science and folk medicine at identifying how to stay healthy as well as why certain practices promote health.

The new study, by itself, doesn't tell us much about what to have for breakfast.

It doesn't tell us how to boost memory and reaction time, because the walnut group didn't consistently outperform the control group (and, when they did, the differences were small – for reaction time, the benefits ranged from 8 to 22 milliseconds, or thousandths of a second. That's less than half the time your eyes are closed during a blink.)

At the same time, the study doesn't explain why the walnut group occasionally showed better (or worse) performance.

I admire the study methods – overall it's one of the best-designed experiments I've seen in years – but the data illustrates how modern science may fall short.

If I want evidence for my personal belief that walnuts boost cognition, I'd be more persuaded by those Kashmiri Valley inhabitants and others who treat walnuts like a sort of "brain tonic".

Modern science does at least confirm that walnuts are healthy. They're an excellent source of protein, vitamins (especially B₁, B₆, and folate), minerals (magnesium, zinc, phosphorous, etc.), and fiber, among other things.

In my view, what's more important than deciding whether to include walnuts with breakfast is to figure out the types and quantities of food that keep you most alert and energized throughout the morning. In other words, noticing some of the micro-associations between breakfast and cognitive functioning.

What works best for me happens to be a cup of strong coffee, a glass of water, minimal carbs other than fruit, and a fair amount of protein (some combination of walnuts, eggs, yogurt, or plant-based meat).

I doubt this menu would be best for everyone – I'm a morning person, for instance, and chronotype is known to affect metabolism. This is just an illustration.

So, should we eat walnuts with breakfast? Sure, but I wouldn't feel pressured to do so.

Thanks for reading!