Are you what you eat?
When it comes to your brain, the answer is a resounding yes.
One brain-friendly food to add to your shopping list: blueberries.
In episode 139, Dr. Robert Krikorian, Professor of Clinical Psychiatry and Director of the Division of Psychology at the University of Cincinnati, talks to Jesse about why you should eat blueberries to improve memory.
The Benefits of Blueberries
There’s been plenty of talk in the media about how blueberries are the new superfood. We’re always hesitant to buy into this kind of hype (remember how margarine used to be “healthy” and eggs were horrible for you?). But in the case of blueberries, the science backs up the hype.
Blueberries improve long-term memory (retention of information over time), access to words and concepts (crucial for dementia or Alzheimer’s sufferers), and short-term memory (aka working memory).
They reverse loss of balance and coordination in older rats.
The benefits aren’t confined to older people, though. One study found that blueberry juice improved memory and concentration in children.
And if you’re worried about consuming too much sugar from fruit, blueberries even lower blood glucose levels.
For more, check out Dr. Krikorian’s latest research on blueberries.
What’s Going On In Your Brain?
Dr. Krikorian stresses that we don’t have a thorough understanding of how blueberries are able to confer all these benefits. “A lot more work needs to be done,” he says.
However, it’s thought that the beneficial effects of blueberries is due to a certain type of flavonoid (the active compounds in blueberries) called anthocyanins. Anthocyanins help plants resist radiation from sunlight and offer a defense against bacterial and fungal infections.
These same compounds interact with our biochemistry in beneficial ways, including helping to improve glucose disposal.
Frozen, Raw, or In Muffins?
Before you nab that blueberry muffin from the coffee shop and think you’re taking care of the old noggin, think again. Not all blueberry products are created equal.
Fresh, whole blueberries are obviously ideal. Dr. Krikorian uses whole fruit in his studies because he doesn’t want to assume that one part of the fruit is what’s actively benefiting memory. But whole frozen fruit is equally good (and often cheaper – yay!), since it tends to be harvested, cleaned, and frozen in about 24 hours.
Two other forms that get the thumbs up:
- 100% pure blueberry juice (as long as there is absolutely nothing else added), although there are some changes in composition from the juicing process.
- Freeze-dried blueberry powder. Benefits: it’s easier to store and consume than juice.
Avoid juices with added sugar or that are made from concentrate, and baked or cooked blueberries, since exposure to heat degrades flavonoids.
A Blueberry a Day Keeps the Doctor Away?
There’s no standard dose of blueberry. Amounts used in research vary widely (some have used one cup of whole fruit per day, others ½ cup).
And you may not need to eat blueberries every day. Certain metabolites persist in your body long after eating the fruit. So, you may be able to get the benefits of blueberries from eating them a few times a week.
Ultimately, more research needs to be done. But in the meantime, throw a handful of blueberries on your yogurt, in your smoothies, and on your cereal.
Other Memory Boosting Foods
Blueberries don’t have a monopoly on boosting memory. Other research has found cognitive enhancement from spinach, strawberries, and grape juice.
Plenty of other blue and purple foods have high levels of anthocyanins, like blackberries, grapes, red cabbage, and eggplant.
|0:22||Blueberries and cognition|
|1:36||This Week in Neuroscience: Updated map of the human brain hailed as a scientific tour de force|
|4:40||The audience interaction section|
|8:18||Jesse’s embarrassing blueberry story|
|9:52||Intro to Dr. Robert Krikorian and his background with blueberries|
|13:22||The state of human and animal blueberry research|
|15:45||Parent compound versus leftover metabolite level effects in the body|
|17:16||Are powdered berries and berry juices as effective?|
|18:22||Frozen, fresh, cooked: does presentation matter?|
|19:56||How much of a cognitive effect can blueberries have?|
|21:57||Memory and semantic access|
|23:05||Acute versus chronic effects|
|24:50||What are flavonoids?|
|27:23||How many blueberries should we eat to see positive effects?|
|28:42||Increased blood flow|
|30:14||Other sources of anthocyanins|
|31:01||“You may be doing yourself some good and probably not any harm”|
|33:35||Ruthless Listener-Retention Gimmick: A robot has been teaching grad students for 5 months... and NONE of them realized|
PS: With your blueberry-powered memory boost, don’t forget to sign up for your weekly Brain Breakfast.
Episode Transcript hideshow
— This Week in Neuroscience --
Jesse: So, scientists at Washington University in St. Louis have recently published a new map of the human brain that has a lot of people in neuroscience very, very excited. So we’ve had maps of the brain for a long time, most of us are pretty familiar with what a brain looks like with the crinkly surface and all that. And for the last 100+ years, they’ve had a variety of different brain maps breaking the brain up into physical structures based on attributes that are visible to the naked eye, like the difference in the folding in the ridges between, say, the cerebellum and the cortex. They’ve made maps based on brain activation, brain connections, and kind of with each new technology there’s been a new mapping. The somewhat frustrating side effect: not all of these maps necessarily correspond with one another that well. It’s like if you imagine a map of the world, most maps of the world are broken up according to nation states, but you could just as easily have a map of the world that’s based on the average heat—there’s not these lines between Canada and the US, and instead everything is just a smooth gradation based on average color. Or you could have one based on topography. You could have all these different things and one might not have much to do with the other. And that’s sort of been the problem, these various mapping systems each probably have their strengths and their weaknesses, but collectively there’s not like a right answer to how should we map the brain.
But this recent study, they did highly detailed MRI studies from 210 healthy young adults, this was part of what’s called the Human Connectome Project, which you can learn more about online at HumanConnectome.org, and what that is is a massive effort trying to understand how the neurons in the brain are connected to one another. If you think of a physical mapping of all the neurons in the brain as being like a phone book, a connectome would be a much more rich piece of information saying for each one of these phone number nodes, here are the other phone numbers it calls, here are the other phone numbers that it’s called by. So this study, this big new brain map has broken the brain into at least 180 different regions, which is quite a bit more than any of the previous mappings. It has confirmed the existence of 83 regions that had been posited by other mapping methods, and has also added 97 new ones. So it’s slicing and dicing a lot more finely, getting very specific in the type of attributes that need to be within certain parameters for that portion of the brain to be considered one unit. Whereas in previous mappings, having a certain type of cells for example, that might qualify for constituting a particular region of the brain, now you would need to have all those same attributes but also react in a similar fashion to the activation of some brain network. So, to go back to the map of the world analogy, you would not necessarily just have nation states, but you would say I want to see a map that divides by nation state and also by topography, so I want to see everything that’s in the United States of America, 3,000 feet above sea level.
They didn’t address this, but I wonder how many new brain area names are we going to need to come up with? We’ve got so many already, and just when I’ve started to remember names like the fusiform gyrus, if they change the name of things like that, I’m going to break down in a corner and cry. But the example that they give of the names in this new mapping are things like 55b, so not terribly exciting or memorable or even scientific-sounding names. I’ll reserve judgement on that one going forward. But overall, it sounds like a great new functional system of classification. Neuroscientists are excited about it, so we probably should be, too.
— Main Interview --
Jesse: So I’ve been mentioning for a couple of episodes now that we were going to have a blueberry episode coming up, and I might have sounded a little unusually excited about this. Everybody’s got kind of their little pet projects and their very favorite things, and blueberries are definitely one of mine. They’ve always been a special food in my food groups; I eat them in embarrassingly large quantities, especially when they’re fresh in the Oregon summertime. My mom used to grow them in a little yard that we had off on the side of the house, and there were big, big fields of blueberry orchards or whatever they’re called, like "U-pick Blueberries," and that was always something we did.
So, I’m a long-time fan of blueberries. There’s a family legend, and I kind of vaguely remember this but I’m not sure if I remember it actually happening or if it was just a story that was retold so many times that I kind of have reconstructed this memory based on having my parents tease me about it, but according to this legend, I was three or four-years-old and I spent so much of the day eating blueberries that I pooped blue the next time I pooped. This was obviously at an age when my mom still interacted with my poop on a regular basis. But anyway, even at an early age, obviously blueberries were something I was very, very into. And it turns out, not to be too disgusting, but the blue in blueberries, the stuff that gives it that dark pigmented color, that’s actually where a lot of the action is as far as the positive benefits of blueberries and other fruits like blackberries, concord grapes, eggplants. A lot of these fruits and vegetables with those really dark colors, they’ve got similar chemical constituents and give us similar benefits. Some of these plants are richer in these anthocyanins than others, and in my highly biased view, kind of rooting for blueberries, I’m happy that blueberries seem to kind of be the winner. But I suspect you could probably make up in quantity for lack of quality by eating some of the other fruits within this category.
So a couple of months ago, we were looking at the calendar, we knew summer was coming, and I asked Michelle on the Smart Drug Smarts staff to go and headhunt us out a blueberry expert, find somebody that knew blueberries inside out, and one name quickly rose to the top of the heap: he is Dr. Robert Krikorian from the University of Cincinnati. For any of you who are attending the Berry Health Benefits Symposium next year, he’s going to be one of the speakers there. He’s been studying berries for the better part of two decades. I just relistened to this interview and it’s a little bit of a softball interview because I wasn’t really going for the hard, incisive questions trying to find out any damning evidence that blueberries might actually have some horrible side effect or something, because I didn’t want to know that. I’m rooting for blueberries to be the awesome thing that I want them to be—which apparently they are. But let’s dig into the science now with Dr. Robert Krikorian.
Jesse: So, to give a little bit of background on yourself, how long has this been going on? How long have you been looking into berries?
Dr. Krikorian: Sure. It’s kind of a personal story for me. I read a paper, the paper was published in 1999 I believe, it was out James Joseph’s lab out of Tufts University in Boston, and then he also was on staff with the USDA facility there. He and his colleagues published a paper indicating that spinach, strawberries, and blueberries seemed to revert motor and cognitive performance in older rats to younger levels even though the supplementation was initiated when the animals were aged in late life. That really captured my imagination, and I followed some of his other work and then I wound up inviting him to come to my institution in Cincinnati to do a grand rounds talk. And so he accepted the invitation, I think he flew in the evening before, stayed in a hotel; I met him in the morning, we had coffee together, and we walked over and he did his talk, which was really exciting, and presented some great data at that point—this was in 2003, I believe. And then he left almost immediately. I mean, I remember right after the talk I was riding down with him in the elevator, he was going to pick up a cab to go back to the airport, and I just said, "Gee, it’s too bad that you’re so far away. It’d be great if we could do some research together," and he just, without skipping a beat, looked at me and said, "You want to write a grant?"
Jesse: Everybody can use a grant writer.
Dr. Krikorian: Yeah, I was taken aback a little bit, but I said yes, sure. It turns out he was an animal researcher and had been passively looking for someone to do the translation to human research for a long time. So I actually, in loose collaboration with him, submitted several grants, and before we got funded, he had a relationship with Welch Foods, the grape juice people, and his lab had done one study with Concord grape juice that showed some benefit in older animals, and they were interested in a human trial, so he put me in touch with their research people, and they gave us just a very, very small amount of money and we were able to do a very small trial with grape juice. And the structure of that trial, I was able to add an arm with blueberry juice, and Jim, again put me in touch with the Wild Blueberry Association of North America, and they provided the juice for that study. And so, what came out of that very small study was a publication showing enhancement in cognitive function with Concord grape juice supplementation in older adults with mild cognitive impairment, which is a risk condition for dementia and Alzheimer’s disease, as well as a small preliminary paper showing similar benefits with blueberry juice supplementation. And that kicked things off; based on that pilot data, a few years later I was able to get funding from NIH to look at blueberry powder supplementation along with omega 3 fatty acid supplementation in a larger trial.
Jesse: Overall, how do you feel about the state of the research where we are now, what we know about things like mechanisms of action, vs. what you’d still like to know? Like, how many of these are human studies vs. animal studies? Sorry, that was like three questions at once. You can pick your favorite.
Dr. Krikorian: You know, the state of the human research relative to the state of the preclinical animal and molecular research with berry fruit supplementation is quite different. There are a growing number of human studies, but much, much fewer than already exist in rodents. But that basic science database is extremely important in terms of promoting the notion of doing more human research and also understanding some of the mechanisms that may be associated with what we’re finding.
Another very important issue is that the level of understanding about mechanism and even about what you might think of as pharmacokinetics in terms of what happens to the presumed active ingredients or active compounds in the berries, the flavonoids, after they’re ingested is very wide open. There’s some new data showing that the metabolites of the anthocyanins, which are the flavonoid compounds in blueberries, persist in the circulation for several days, to several weeks, and potentially months—we don’t really know for how long. The parent compounds, the parent anthocyanins, don’t stick around very long at all, they’re gone within a matter of hours. And so before more recent data came out, a notion was that the parent compounds were distributed throughout the body and into the brain and other places in the body, but they were not measurable after maybe four to six hours or so. And so, my thought given that was well maybe what’s happening is its signals are being switched on or off by the ingestion of the parent compounds and they were having longer-term effects. But now we know that the metabolism, that there are 400+ metabolites of these compounds that are potentially being broken down and the reconstituted as they circulate between the liver and digestive tract over a several-week period, as I mentioned. And so, it’s possible that that process, that the body processing the compounds, has an effect as well that we don’t understand.
Jesse: That’s interesting, that’s almost a direct parallel to a conversation I had just a few days ago about the compound ketamine, which is an anesthetic, which is also being used now for some antidepressant potential work. And the same thing, it seems like the metabolites that come out of ketamine are persisting in the body far longer than the ketamine itself and might actually be responsible for some of the mechanisms of action seen there.
Dr. Krikorian: I mean, this is preliminary data that’s actually unpublished and I’m actually working on writing it up right now, is in our most recent studies with blueberries in older adults, we’re seeing that the levels of metabolites in circulation in both the placebo group and the active supplement group don’t change very much over the period of the study or the period of the intervention. But the level of the parent compounds, the anthocyanin parents, do change, are increased in the supplemented group. So to my mind, that begs the question, well, is it still really just the parent compounds that are producing the benefits because we have been able to measure some cognitive benefits and some changes in brain activation using brain imaging in the blueberry supplemented group that we don’t see in the placebo group even though the metabolite levels seem to be about the same? So to my mind, that suggests that it’s, again, the parent compounds. But it doesn’t rule out the possibility that there’s other mechanisms at work. And if the latter is true, if there’s a hormetic effect or some effect associated with the metabolism of the compounds, that’s happening within the placebo group AND in the active supplement group. One might infer from that that it’s hard to measure a benefit if really both groups are getting it. Because we’re looking at diet records in these recent studies as well, and we prescribe for both groups to not consume berry products or berry supplements of any sort in their diet during the period of the study, but they still get flavonoids, including anthocyanins, from non-berry foods, and so it’s a complex picture.
Jesse: So you mentioned powdered berries there and also doing studies with juices. Are those two ways of getting berries functionally identical, and can you tell us something about the juice? I mean, I know that at the really bad end of the juice spectrum you can buy juice made from concentrate that’s like 10% original juice and 90% water and high fructose corn syrup. I’m assuming that sort of Kool-Aid-style juice isn’t involved, but what are you doing these tests with?
Dr. Krikorian: Yes, both the blueberry juice and the grape juice that we use is 100% fruit juice with nothing added. You’re correct in the sense that there are some changes because of the processing in the compounds of interest, the flavonoid compounds, but that’s something that can be assessed—the composition, the flavonoid components in the juice. We’ve moved to powder, which is a whole fruit freeze-dried powder, because it’s easier to work with. Some of our participants had purple trunks when they’d take home several bottles of juice for the study and sometimes some of the bottles would break. But it’s much easier to use powder, it’s a fairly good representation of the whole fruit, and it’s easier to store and consume, and we can package it in individual dose packages, so we have better control over the dosing as well.
Jesse: So, total self-serving question here, but from your perspective nutritionally, is there a difference that people should be aware of between raw blueberries, cooked blueberries, blueberries that have been frozen and then thawed out, anything like that?
Dr. Krikorian: Again, there have been some measurements of the flavonoid composition or contents after different kinds of processing. It’s a little bit outside the scope of what I do, but to my knowledge based on some conversations with processors and some of the people that measure these things, my sense is that there’s not very much difference between whole frozen fruit that’s not been otherwise processed because it tends to be harvested, cleaned, and frozen within about 24 hours maximum, sometimes even much less than that, and the preservation of the content is pretty good, I think. Now, if you thawed out with exposure to higher heat levels, you’d have the possibility of changing the composition somewhat, and so that’s a consideration. But fresh fruit and quickly frozen whole fruit is roughly equivalent, as far as I know.
When it’s baked, time, exposure to moisture and exposure to heat are factors that will cause change and dissipation in the flavonoids. So I think if you’re getting blueberries in a muffin or in a bakery product or something like that, it’s likely that there’s change because of exposure to heat, and you may be getting less of the good stuff, so to speak. Although, there’s heat exposure in juice and that can also stimulate recomposition of some of these compounds, and that may be positive, too. Again, the word is complexity. This is a very complex field and we don’t really understand what is producing the benefits or even how these compounds should be processed to maximize—or not processed.
Jesse: So for the human studies in particular, they’ve achieved statistical significance, so odds are good that they’re actually doing something. But what is the size of the impact? Is this making a large difference to cognition? A small difference? What sort of "oomph" are we talking?
Dr. Krikorian: I would say moderate, in the middle. Statistically, the effect size is, in the work that we’ve done here, for some of them actually were large, and they would have had to be because we had so few—the sample sizes were about 10 per group, so quite small. Yeah, they’re still considered pilot studies, so in some of those studies we had a large effect size. But I generally expect what are considered moderate effect sizes. I would count those as meaningful in terms of influencing functional capability, everyday function, that sort of thing. The percentage of effect varies quite a lot and it’s not exactly the same as an effect size, but sometimes these can be as high as, say, 50% or 40%, sometimes even a 60% or higher change relative to the placebo group’s performance on a task. But then there’s variability that has to be taken into account; variability among the participants within a group.
One of the messages that I think one should take from this is that we’re still very early in the human studies, a lot more work needs to be done, and we think we have some ideas about what are the compounds that are associated with the benefit, but we don’t really know, and one example in that regard would be fiber. I’ve chosen, in my studies, to use the whole fruit powder in these second generation studies, which I think is a good idea because I don’t want to make the assumption that a certain cluster of compounds are where the action is and maybe leave out something that should be included to get an effect. And fiber is potentially a bioactive substance as well, or it is a bioactive substance; fiber has a lot of beneficial effects, and there may be interactions, with the parent anthocyanins for example, that are very important. So I also chose, in these studies where we’re using powder, to not put fiber into the placebo powder because I thought if there was a benefit from the fiber, I didn’t want that to lower our chances of finding a signal if there was one to find.
Jesse: For the actual cognitive changes that you’re seeing, do they cluster in any sort of area? Are these benefits to memory, or motor control, or executive function? What’s on the scorecard?
Dr. Krikorian: We’ve seen performance changes in memory—that is long-term memory, which is operationalized in these studies as exposure to some information and then retention over time. That’s been something that’s shown up in a few studies now. We’ve also see some benefits, actually in one of the most recent studies, in semantic access, which is an important function if you’re thinking about Alzheimer’s disease. Decline in semantic knowledge, which is the store of basic conceptual information, like who was the first president of the United States, general knowledge, general conceptual information, that tends to decline in Alzheimer’s disease, and in pre-Alzheimer’s as well. We had a finding in one of the more recent studies showing improvement in the supplemented group in access to that kind of information. Also, the suggestion of some improvement in working memory, although that hasn’t been as consistent an effect as the long-term memory improvement. And when I say working memory, I’m thinking of working memory as a component of executive ability.
Jesse: Have these effects shown up as acute effects, like you eat a handful of blueberries or a bowlful of blueberries and then they’re there for a few hours and they fade away? Or is this sort of an ongoing thing that will stay with a blueberry user for a long time to come?
Dr. Krikorian: Well, the design of our studies is to do an intervention over a period of several weeks. We’ve done 12 weeks, 16 weeks, and even 24 weeks in one study. So we do cognitive performance measures just before the supplementation is initiated and then during the final week of the supplementation. So we don’t sample very much effects after a week, say, or after a day. There are some other studies, some of the British studies have looked at cognitive performance over the course of hours and shown some benefits, lack of fatigue, or lack of deterioration of performance I should say, over a course of hours with blueberry supplementation. And there’s some other data that’s not cognitive data but brain activation data using different kinds of flavonoids—flavonols, which are contained in cocoa powder for example, and also they’re represented in Concord grape juice—and looking at brain activation and enhancement of brain activation occurs acutely based on what we know so far. And so, my sense is some of these effects are happening very quickly, potentially within hours, but then also chronic consumption over time seems to be showing benefit as well. In a couple of our studies, we’ve looked at brain activation using fMRI in people consuming Concord grape juice and also whole fruit blueberry powder and we’ve seen some increased activation in the grape juice and blueberry supplemented individuals. We’re getting performance differences on cognitive capability, but now we’re moving a little bit into getting evidence of some change in brain activity, which presumably is beneficial as well.
Jesse: You sited flavonoids as being the likely candidates being responsible for a lot of these positive effects. What do we know about what flavonoids do in the plants?
Dr. Krikorian: Generally speaking, polyphenols of flavonoids are not associated with reproductive activities of the plants, they’re more associated with defenses. So, the anthocyanins are these pigmented compounds, the deep blues and purple colors that you see with blueberries, for example—or even purple cabbage, those are anthocyanins as well—and they help the plant resist exposure to high levels of radiation from sunlight, defenses against bacterial and fungal infections, and perhaps other kinds of defensive actions as well. It turns out the compounds in these plant chemicals interact with the mammalian biochemistry in different ways and seem to produce benefits.
Jesse: A lot of the time it seems like when we find a plant that has positive effects on human physiology, it turns out that the active compound in the plant is some sort of close analog biologically to something that we actually have within our systems. Is there anything like that with the flavonoids, where there seems to be a close endogenous analog within the human body?
Dr. Krikorian: Not that I’m aware of. There’s recent data—when I say recent, probably within the last five years or so especially—suggesting that, for example, anthocyanins are beneficial with respect to glucose disposal; they may be beneficial with respect to helping to prevent type 2 diabetes, that’s been mentioned by some other researchers. Some of the data that I’ve looked at seem to indicate that the anthocyanins help to improve glucose disposal regardless of the level of endogenous insulin, which suggests that in some sense maybe they’re performing the same action as insulin or, in some other indirect way, improving the uptake of the glucose into the tissue.
Jesse: That is just about the best news that you could’ve possibly given me, because the one thing that I worry about, I love fruit, but there is a lot of sugar there. But if there’s something in blueberries that’s sort of helping to manage its own sugar, that is really, really good news.
Dr. Krikorian: Well, there’s some really nice acute studies done by another colleague of mine who’s in Chicago, she’s done a number of studies with younger and middle-aged individuals who are basically healthy and they’re fed a high glycemic/high sugar/high fat meal with and without strawberry powder, and the figures are extraordinary. I mean, you see the people that don’t receive the strawberry powder show the typical spike in glucose and insulin after the meal extended over a few hours, but it’s very much flattened if you add that strawberry powder, if it’s consumed with the meal. Quite striking data.
Jesse: So, thinking about the strawberry powder in that study or how many blueberries to buy when I go to the store, to pick up some of these benefits that we’ve been talking about, how should one be titrating one’s blueberry consumption? Are we talking handfuls, or buckets, or barrels, somewhere in between?
Dr. Krikorian: Well, the answer to that question is I don’t know, but I can give you sort of an educated guess, and I think it’s less than we thought. In the first generation studies, we were positing roughly the equivalent of one cup whole fruit per day, and in the studies that we’re planning to do now, we’re dropping that down to one half-cup equivalent of a whole fruit per day. And given the data from the lab that I mentioned earlier, about the persistence of the metabolites and so far, I wonder whether it doesn’t even have to be daily supplementation, it could be perhaps two times a week or four times a week. But it may not be a matter of buying a pound of blueberries a day and setting yourself to eating them every single day. I think we can take smaller amounts somewhat less frequently and still obtain the benefit. But again, that’s not really known; I’m making an educated guess about that.
Jesse: See, I’m the sort of blueberry fiend that if you’d just said, "Hey, you should eat a pound a day," I would’ve taken you at your word and happily done it on good faith.
Dr. Krikorian: Yeah, there is some evidence that there’s a point of saturation, maybe that at some point some of the metabolic mechanism are overloaded and more gets diverted to excretion or elimination from the body than at lower levels.
Jesse: Do we have much of a sense, for the cognitive effects from blueberries, what’s actually going on to cause those changes, or are these correlational studies that just aren’t going to get that sort of information?
Dr. Krikorian: I think generally, certainly in humans, it’s more correlational. In the studies we’ve done with brain imaging, what we’re actually measuring is blood flow-related phenomenon. But blood flow is very highly correlated with regional brain activity. So if a certain part of the brain is more activated, there are mechanisms to divert more blood, supply more nutrients and oxygen to that part of the brain—that’s the basis for fMRI. So the question would be well what’s happening when an individual is consuming, blueberries for example, on a regular basis? It may be part of what’s happening is an improvement in the circulation, more blood is available. And flavenols, the cocoa flavonoid, I mentioned they’re also represented relatively well in Concord grapes, have been studied for blood pressure reduction in people with hypertension for many years and there seems to be some sort of benefit in that regard. And the mechanism has been worked out, which seems to be an increase in the synthesis of nitric oxide, which has a relaxing effect on the blood pressure walls, which allows more blood flow, that may be one mechanism for the change in activation we’re seeing. But then you’d make the speculation perhaps that if you’ve got more blood flow available especially in older adults, where you would predict a reduction in general of blood flow, then you may get better delivery of essential nutrients and oxygen to brain centers and those centers may function better, and that might be at least part of the basis for the cognitive benefit.
Jesse: You mentioned Concord grapes in addition to blueberries, and also purple cabbage. Are there any other plants that might be listed alongside these as sources of anthocyanins?
Dr. Krikorian: Well, strawberries contain anthocyanins; there’s purple corn, I guess, that does as well.
Jesse: Oh, yeah.
Dr. Krikorian: Anything that’s purple potentially has anthocyanins. But to my knowledge, berries might be considered the primary source. When we enroll volunteers in our studies, we give them what’s called a forbidden foods list that we’ve developed, and mention in that list that they should not consume any berry fruit, any berry drink, any berry extract or supplement for the period of the study because we’re trying to look at the effect of the supplement we’re providing.
Jesse: What about eggplant? Is eggplant on that list?
Dr. Krikorian: Yeah, actually that’s another good option.
Jesse: I don’t know why anybody would choose an eggplant over a blueberry, but maybe there’s somebody out there. Anything else? Any final thoughts as people plan for their seasonal summertime shopping?
Dr. Krikorian: The two things I would say is it’s too early to know with respect to a lot of these interesting questions about mechanism and dosage and frequency of dosage. There have been just a few studies and we’re doing more, but many, many more need to be done. The other point to make is even though all the questions aren’t answered, I think there’s enough information now to believe there’s a probability or a likelihood that there’s benefit from consumption of blueberries and other berry fruits, and probably little or no negative side effect unless, you know, you can overdose on anything, obviously. But taken in moderation or sensible amounts, you may be doing yourself some good and probably not any harm.
Jesse: So, thank you so very, very much to Dr. Robert Krikorian for taking the time for that interview and for helping to push the science of blueberry consumption to the next level. It’s one of those health maxims that you sometimes hear when people are talking about how to have a proper diet, is to eat lots of vegetables with lots of different colors. And at first blush, the first couple of times I heard that I thought, "That’s just gotta be some kind of cute little super-simple heuristic that oversimplifies but is easy for people to remember, and so they’ll just say it that way even though it’s kind of skipping over the science. But kind of counterintuitively, the more I talk with people about plant-based nutrition, it really does seem like the colors map pretty closely to what’s actually going on as far as the chemicals within different plants. And sort of that dark blue, purple-y cyan color, that does have a certain chemical profile. Yellow vegetables, like some peppers and things like that, they have a different chemical profile. It’s probably not exactly a one-to-one correspondence between different chemicals and all the colors of Crayola crayons, but there really is something to be said for making sure that you have a variety of different colors on your plate without resorting to food coloring or Skittles or things like that.
And actually, just to wax on nutrition a little bit, it seems like oftentimes the biggest improvements that people can make to their diet is actually getting certain bad things out of their diet and sort of substituting in something else that gives them the calories, fills them up, but has a better nutritional profile. And when things are sweet and tasty, and I think most people would probably agree that blueberries fall into that category, that potentially gives you a pretty good opportunity to swap out something that might also fall into that sort of sweets category that isn’t nearly as good for you. Substitute in something that kind of tastes parallel, has that sugar, but also has all these health benefits, and kind of get the dietary double-win doing something like that. So anyway, everybody enjoy your summer berry season, and remember what Dr. Krikorian said, that fresh frozen fruits are pretty much nutritionally identical to fresh fruits. So, this doesn’t need to be a summer-only thing, this kind of stuff can go on all year-round. But now let us move ahead to the Ruthless Listener-Retention Gimmick.
— Ruthless Listener-Retention Gimmick --
Jesse: So it seems like one of the main cultural dialogs that we’re having right now involves a lot of pushback against the "technolog-ification" of society—I’m not sure if that’s the right word or a word at all. But in the 1950s/1960s, there was this concept of better living through chemistry, they just started making massive progress and inventing new plastics and stuff like that, and inventing foods like Twinkies that you really couldn’t make out of a pantry of food products without resorting to some real weird chemical test tube stuff. And the idea kind of naively back then was that anything invented is automatically going to be cool, and of course we’ve since learned that there’s a lot of downsides to man-made inventions that we aren’t necessarily equipped to handle well, whether that’s digesting Twinkies en masse or breathing the fumes that come out of our labor-saving devices, like cars or one of my most hated technology objects, leaf blowers. I can’t believe that people use leaf blowers instead of rakes. They’re loud, they’re not that fast, they’re smelly and just, god, I hate f****** leaf blowers. And so you’ve got kind of these back-to-nature movements and the paleo diet. I eat the paleo diet myself, so I’m certainly not knocking some of the blowback against the "technolog-ification" of everything.
But one of the things that people ask me about a lot, and this includes family members that are like, "Jesse, you might have a problem," is the amount of human-to-human interaction that we have now that is digitally-mediated somehow, that’s not me sitting across the table from you or taking a walk with you but is me and a friend having a conversation on Skype or exchanging emails or a chat session with my girlfriend, or one of these things where there’s a computer in between two humans that are ostensibly having, growing, or are maintaining a relationship. I’m open to being wrong here, but I kind of tend to think that there’s a little bit of "sky is falling" about this. I’m certainly not knocking direct human-to-human, in the same room, physically localized interactions, but I kind of think something like that is like vitamin C, where you need a minimum amount of direct human-to-human interaction, and once you’ve kind of reached that amount, then having a bunch more probably isn’t going to hurt you but you don’t really need it, either. I’ve maintained for a long time that I feel like the two things that I need people in my physical proximity for are sex and sports, and aside from those two things, I’m happy to do almost 90% of my relationship stuff—I don’t mean romantic relationships, I just mean relationships in general—through some digitally-mediated platform. That’s me. I’m not saying I’m right, I’m not saying I’m wrong. There might be some really interesting studies that show that that’s just a horrible idea, but I haven’t seen them yet; I’ve yet to be persuaded to the contrary.
But one really interesting aspect of these digitally-mediated communications is that you’re never really sure who you’re dealing with. Is there somebody eavesdropping on your email, or is the person that you’re emailing the person that you think you’re emailing? We’ve all heard the stories, maybe they’re apocryphal, maybe they’re not. Certainly seen the internet memes about somebody going on to an online dating site and thinking that they’ve found the boy or the girl of their dreams and it’s actually somebody that’s far from the person of their dreams—a morbidly obese person living on their mom’s couch, pretending to be some sultry young 20-something. And while that’s on the extreme end, if this hasn’t happened to you, it’s happened to somebody that you know, has gone on a date with somebody that started as an online interaction and the person in real life didn’t really match up with the photos that were shown online. Maybe there was some photoshop trickery, maybe it was just a very, very careful selection, but the real world doesn’t always bear up with what we were led to expect we’re dealing with in an online scenario.
And so it was with students at Georgia Institute of Technology, which offered an online course in artificial intelligence. Lots of students were taking this, there were a total of 10 TAs that were supporting the students—TAs are teaching assistants. The students would read their coursework and do their assignments, and if they had questions or wanted advice on something or another, they would email into the TAs and get their questions answered. One of the most beloved TAs, this was a class starting in January of 2016, the gig was up by mid-May… One of these TAs by the name of Jill Watson was, in fact, an artificial intelligence construct. If you remember the computer by IBM named Watson that won big on Jeopardy a couple years ago, well, Jill Watson was powered by the very same Watson by IBM and she was programmed full of information so she could be an effective teacher’s aid in the class on artificial intelligence. She would answer questions by email, they set up a fake Facebook account, a fake LinkedIn account for her.
So as far as the students could tell—and none of them really suspected anything until pretty deep into the course when one guy kind of got wise to this and was like, "I wonder if an AI is hiding around here somewhere since this is a class on AI." So it wasn’t necessarily that he suspected her in particular, because he originally asked the professor, "Hey, are you a real human?" But in fact it was Jill Watson, this quick-on-the-draw with her email responses AI that was just waiting there with appropriate answers to hundreds and thousands of student emails and interactions. One student who was fooled said, "It seemed very much like a normal conversation with a human being," and another student said, "Just when I wanted to nominate Jill Watson as an outstanding TA…" I don’t know why he still couldn’t—she was doing her job.
Some people have subsequently questioned the ethics of hiding an artificial intelligence entity somewhere where they can be interacting with humans who aren’t tipped off to their status of the AIs being an AI. But defenders of Jill Watson’s appointment to this post point out that finding good qualified competent TAs is not easy, and a TA that can work 24 hours a day and instantly respond to emails is worth her weight in gold or Silicon wafers or whatever it might be. They configured Jill so that she would only answer questions when she had a 97% likelihood that her answer was going to be correct, which is a pretty firm threshold. I mean, I answer questions when I feel much less than 97% correct, so I think that’s a pretty legitimately high standard to hold one’s self to. And the way that she was able to do these internal calculations, say, "How sure do I feel about this?" Here I am calling her a "her" rather than an "it"… But was to gather the incoming data, the email that the student had posed a question in, and then approach the problem in a few different ways, look at different ways of potentially answering that question, and if the different approaches yielded the same answer, then that was a pretty good clue that this actually is the right answer. If she tried several different ways of attacking a problem and came back with lots of different answers, then she’s not so sure.
So, interesting the ways that approximating human intelligence through AI is not necessarily going through equivalent mental processes. I don’t know anybody short of maybe the world’s most rigorous scientist that would actually approach a regular problem that way, trying to solve it several different ways and seeing if the answers coincide. But it reminds me of something that Dr. Ken Ford said on an AI episode that we did a long time ago, which was making an analogy to artificial flight and saying how artificial flight went no where as long as people were trying to mimic the way that birds flew, and if we stuck with that approach, we probably still wouldn’t be airborne now. But the key to solving the problem of artificial flight was kind of going back to the drawing board and coming up with an all-new way of doing it. It may well be that under the hood there’s not a whole lot of correlation between the way that a human brain works and the way that our best AI systems might be working now and in the future.