How Much Fruit Is Too Much? A Scientific Look at Fruit Like Never Before

Is fruit healthy? And how much fruit is too much fruit? These aren't easy questions to answer, but in this video, we will weigh in on this debate and dissect these questions with the critical thought they deserve in a manner that hopefully you'll appreciate. Okay, I promise, no more puns. To give you a roadmap, we're going to start by discussing fructose and explain why the fructose in fruit is contextually different than the fructose in many processed foods like ultra-processed candies or sodas. And we will then review why the scientific literature on this matter, on the topic of fruit, can actually be very misleading, going through limitations of the literature broadly. Finally, we will end with a practical guide helping you to rank fruits based on their various parameters. By the end of the video, I think you'll have a different perspective on fruits than you've ever had before. But let's start. Part 1: Fructose vs. fruit. Now, fructose as a molecule is uniquely metabolically harmful. It acts in the liver to promote metabolic syndrome, it depletes cells of energy and harms mitochondria, and it potentially can even enhance the growth of certain cancers. These are all topics I've dived into before on this channel, and I'll link more references below if you want to go down the rabbit hole. But does this mean that fruit is bad? Perhaps counterintuitively, no. And here's why. The negative effects of fructose depend on its metabolism in the body, starting with the liver as the first-pass metabolizer after, but after absorption in the intestines. However, and this is the tricky bit, the intestines have a way to transform fructose into other molecules, including glucose and organic acids before the fructose gets to the liver. So let's look at some data together. What you're looking at here is what happens when researchers give animals an equal combination of glucose and fructose. However, in any given experiment, either fructose or glucose is labeled, and what you're looking at is the concentrations of glucose or fructose in the portal vein leading from the intestines to the liver. And what you see is glucose passes straight through the intestines en route to the liver, but fructose does not. It gets transformed, leaving actually very little fructose going to the liver. And if you're curious, here's a breakdown of some of the specific molecules into which fructose is transformed by the intestines before it gets to the liver. However, and here's the catch, higher doses of fructose, as in sugar-sweetened beverages, can be harmful since they exceed the capacity of the body, of the intestines, to filter and transform the fructose, basically leading to fructose spillover onto the liver at higher doses. In fact, the authors of a seminal paper, this seminal paper, on the topic of fructose metabolism write, "The balance between fructose consumption and intestinal fructose clearance capacity determines liver exposure to dietary fructose and thereby fructose toxicity. So the next big question, the practical question, is which dose of fructose is too much? Now, based on these data, the upper bound is probably around 1 gram of fructose per kilogram of body weight, which is quite a high threshold. For me, that would be the equivalent of about 2 full mangoes, which are the highest fructose fruit. However, that number is based on mouse data, and given that mice have faster metabolisms relative to their body size, I'd expect a lower threshold in humans. We'll get to the limitations and why these data weren't collected in humans later. And what's more, if you look at the curves from the Keys study, you'll note that fructose in the portal vein, which means it's already spilled over past the intestines and route to the liver, starts to rise around 0.25 grams of fructose per kilogram, even in mice. Now the Nick equivalent dose of that is half a mango. But it gets even more complex since intestinal fructose processing and clearance is also influenced by prior exposure to sugar and fructose. So in reality, since I, using myself as an example, eat a very low-carb ketogenic diet and very little fructose at baseline, my threshold for fructose is likely much lower. By contrast, a fruititarian, I suspect their intestines will be adapted to a higher fructose tolerance. So combining all this together and coming up with rough numbers for you, I would suggest a daily threshold of 0.5 grams of fructose per kilogram for someone eating a standard mixed diet and 0.25 grams of fructose per kilogram at a pretty liberal level for someone on a low-carb diet who doesn't normally eat sugar or fructose. Now, I'll caveat heavily that these numbers I just made up are extrapolations from the available mouse data. You can't do these experiments easily in healthy humans for ethical reasons. And of course, there are many other dietary factors that can interact with fructose intake. So take these numbers with a grain of salt. They're really just my best guesses, extrapolations. And in case you're wondering, for the nuanced ninjas out there, high fructose corn syrup and sucrose, which is made up of glucose and fructose, are metabolically equivalent in this respect, with respect to fructose spillover onto the liver because they contain similar proportions of fructose. The authors actually write, since they did this study, "Sucrose and free fructose, as in high fructose corn syrup, are metabolically equivalent with respect to the processes we just described." Okay, hopefully that makes sense. I know it was pretty nuanced. But moving on, part 2: why is the literature misleading on this topic? It's actually surprisingly difficult to assess the impact of fruits on health. This is largely due to the fact that in large-scale epidemiological studies, there is healthy user bias, where those who eat more fruit tend to live overall healthier lifestyles. Additionally, foods can get lumped together, such that a serving of raisins, for example, and a serving of organic blueberries are counted as similar, which can further bias the data, even though raisins and blueberries are not metabolic equivalents. And in shorter-term randomized controlled trials, where they do exist, one needs to ask whether short-term effects, on biomarkers actually translate into long-term benefits on health. Also, we need to ask whether results in one population translate into another different population. For example, an athlete or generally healthy person having a banana before a run or after a run is very different than a person with diabetes chopping a banana into their morning brown sugar oatmeal. Context matters. And finally, we need to consider what is the comparator? If the comparison is a standard American diet, and increasing fresh fruit intake displaces fruit juice, candy, donuts, then yeah, we can of course expect a benefit or to see health benefits of fruit displacing these other foods. But that doesn't mean fruit intake is metabolically optimal in all circumstances, nor does it mean that a bowl of fruit is a healthier breakfast than eggs and bacon, or a healthier dessert than cheese for that matter. For these reasons, the scientific literature on fruit can be manipulated to reinforce particular narratives, status quo narratives, not because the data are actually watertight, but because of the practicalities and deficiencies of the literature that hinder us from asking certain questions. So, by way of example, we could ask the question, does eating a very high mango diet increase risk of a certain cancer? As compared to eating a ketogenic diet rich in isocaloric amounts of bacon. As much as I'd love to see this randomized controlled trial, it's not happening, so we need to interpret the literature based on practical limitations of the data we can collect. I'm sorry that's complicated, it's just the truth. But moving on, part 3: A Practical Fruit Guide. Although fruit can be ranked along many different parameters, and no single parameter gives the whole picture, For the sake of the exercise, here are fruits ranked in terms of fructose content, glycemic index, and fiber content. Of note, these rankings will of course shift depending on the units you use, whether they're ranked per calorie, per serving, per whole fruit, etc. And you could also find more details and numbers linked in the newsletter below in the video notes. But anyway, the highest fructose fruit is mango. With a full mango having about 30 grams of fructose, depending on the cultivar. Other high fructose fruits include jackfruit, grapes, and watermelon. Lower fructose fruits are berries, with raspberries, for example, having only 3 grams of fructose per cup, and kiwi and citrus fruits, all pretty low in fructose. And of course, the non-sweet low-carb fruits like coconut, avocado, and olives. And if you're wondering where honey falls, it's about 8.6 grams per tablespoon. Now, moving on to glycemic index. The medium to high glycemic index fruits are watermelon with a glycemic index of about 75, meaning it's going to spike your blood sugar a lot. Pineapple at 66, mango at around 56, and banana between 45 and 60. Basically any sweet tropical fruit will have a higher glycemic index. And most dried fruit is a sugar bomb that is typically less than ideal metabolically speaking. Now, lower glycemic index fruits are berries at about 25 and grapefruit at about 22. But two caveats here I want to present. First, the amount of fruit you eat is super relevant. Simply eating half a banana is not the same as eating 3 bananas. No duh, right? That's the difference between glycemic index, which is independent of portion size, and glycemic load, which is the glycemic index multiplied by the portion size. So it really does depend how much of a given fruit you eat. Second, and maybe more importantly, these numbers are based on populations, and we know that individual differences in microbiomes can cause people to have very different responses than population norms. So for example, drawing data from one study, A person might spike in response to a cookie, but not a banana. And another person might have a glycemic response that is opposite, spiking to the banana, but not the cookie. And the same applies to fruits. Pretty weird, right? And in terms of fiber, the higher fiber fruits— well, after the king of high fiber fruits, which is the avocado at about 10 grams per fruit, include wild berries and pomegranates, also with about 10 grams per 4-inch diameter fruit. And basically anything where you're eating the skin is gonna provide some decent amount of fiber, with grapes being excluded. Sorry, grapes. And as an aside, banana peels—yes, banana peels—are about 80% fiber, and I've known quite a few people to actually freeze banana peels with a slather of honey on it. I kind of want to hate on it from a taste perspective, but honestly, I did try it and it's not terrible taste-wise. Although I'm not necessarily recommending it, especially from a fructose dosing perspective. And of course, different fruits have different vitamins and mineral profiles. For example, citrus fruits are rich in vitamin C, along with strawberries, with 8 large strawberries having 100% of your RDA of vitamin C, and kiwis, with 1 kiwi having about 85% of your RDA of vitamin C. But a big note of caution: be careful when Googling high vitamin X food or high vitamin X fruit. Is you can get profoundly misleading results. For example, Google will tell you that a mango is rich in vitamin A, but 1 cup of mango, which is 165 grams, contains only 89 micrograms of vitamin A, or vitamin A precursor, which is only 10% of your daily value. By comparison, just 1 single ounce, 28 grams, of beef liver contains over 2,000 micrograms of vitamin A. So while mango might be a relatively high vitamin A fruit, it is not actually a high vitamin A food. Or as another example, take this joke of a tweet from WebMD, which has 3 million Twitter followers, claiming a banana is a good source of calcium. One large banana has about 7 milligrams of calcium. So you need to eat 150 large bananas per day to get your recommended intake of calcium. Bananas are not a good source of calcium, irrespective of what WebMD says. So if you're interested in the nutrient of a food, what I recommend is Googling the absolute amount and then cross-referencing that to the recommended daily allowance or recommended daily intake where applicable. Now for a few special fruit fun facts for entertainment purposes. Strawberries are not technically berries because they have more than one ovary, They're actually aggregate fruits. Botanically speaking, a berry is a fleshy fruit that has many seeds inside. So a watermelon is actually a berry, but a strawberry is not. That's kind of funny, right? Another fact: bananas are clones. Yes, they're all clones of each other. And here's one: apples in most grocery stores can be around a year old when you buy them. It's kind of freaky, right? And this is enabled by picking them at a slightly underripe stage and then treating them with a compound called 1-methylcyclopropene. Or 1 MCP for short, waxing them, and then reducing the oxygen environment and increasing the carbon dioxide in storage rooms to slow down the ripening process. Kind of freaky. Here's another one: olive trees can live over 1,500 years. For example, the Olive Tree of Vouves— I think I'm pronouncing that right— is over 2,000 years old and is still producing olives to this day. Now what I want you to do is Google another fun fact and leave it in the comments for the community. Let's see which of your comments can get the most likes. But moving on from fun facts, and I hope you do share them, in summary, we can reconcile the fact that fructose can be metabolically harmful beyond its calorie load and that fruit might not be harmful when we consider how fructose is transformed in the small intestine before it hits the liver. Yes, dose does matter, and fructose from blueberries is not the same as fructose from soda. Not from the liver's perspective. Okay, fruit can be included as part of a healthy diet, but it's also not necessary for a healthy diet. And in choosing what fruits are best for you, consider what variables matter to you—for example, the fructose content, the glycemic index or glycemic load, the presence of certain nutrients—while keeping in mind the big picture point that no one single parameter captures the impact of a whole food. That is a general truth. And finally, you can just apply the n=1 approach. By this I mean, observe how eating a food or fruit makes you feel and what it does to your biomarkers in as controlled a way as possible. Because in the end, your n=1 life matters most to you as an individual. And if you choose not to eat fruit and you're super healthy, I'm happy for you. And if you eat a ton of fruit and you're healthy, I'm also happy for you. I think we can all celebrate that. We just want metabolic health to be mainstream. That can mean fruit or it cannot mean fruit. Anyway, stay curious. I hope you found this useful.