For the past year, I’ve been tinkering with an idea—an idea to make a fart-free bean.
Get it out of your system now; this article is about farting. More specifically, it’s about beans, the uncomfortable bubbles of gas they create in our digestive tracts, and whether we can do anything about it. It's not a new question,* as the countless articles that have been churned out over the years attempting to answer it make clear. But those articles don't do much more than repeat the same old advice, much of the time uncritically and with little to no evidentiary basis. Unwilling to pass along the same folk medicine that so many other publications seem to think is sufficient, we decided it was time to do our own tests. Is a fart-free bean possible? Science to the rect...um...um—to the rescue!
*It's also not a frivolous one: Harold McGee, the foremost author on the science of food in the kitchen, ended up writing his seminal work, On Food and Cooking, because someone asked him why beans make you fart. If that’s not enough for you, also note that St. Augustine (who knew of people who could discharge their farts in odorless melodies), Montaigne, and Ben Franklin all expounded on gas (primarily to tell you not to worry about it, advice we are ignoring). And it is said that philosopher-mathematician Pythagoras (who, PS, did not invent the Pythagorean theorem, but that’s a tale for another day) was allegedly afraid of both beans and their magical toots.
Wanting to do a scientific study on beans and intestinal gas is one thing, actually doing it is another. I didn’t have a way to conduct an extensive flatus survey, and I didn’t have access to the fancy laboratory equipment needed to analyze bean samples objectively. I wanted to test a whole passel of gas-reduction theories, which made my problem even more difficult. I was about to give up on the project entirely when a way forward revealed itself. Harold McGee and I have been lecturing together at Harvard’s Science of Cooking class every year for the past decade. At this year’s session, I loudly moaned about my inability to tackle the fartless bean problem scientifically. To my surprise and delight, professors Pia Sörensen and Dave Weitz, who run the class, agreed to measure bean fart potential as part of a class project. They had access to all the necessary equipment—centrifuges, liquid chromatographs, mass spectrometers, and freeze dryers. Even better, they were willing to recruit student volunteers to assemble what we couldn't help but dub the "Harvard Fart Squad," a group of the nation’s best and brightest young minds who would put their GI systems on the line for self-reported, in-vivo science.
The question was, what to test?
What Causes Bean-Induced Flatulence?
Before one can decide what to test, we need to know what we're looking for. Most toot-inducing foods contain ingredients called FODMAPs (an acronym for Fermentable Oligosaccharides, Disaccharides, Monosaccharides, and Polyols—quite a mouthful). Humans cannot digest FODMAPs, but bacteria in our gut can. The byproduct of that bacterial digestion is gas. The FODMAPs that make beans the musical fruit are a small group of oligosaccharides (complex sugars). In theory, if you can reduce the amount of those oligosaccharides in your beans, you can reduce the amount of gas they produce.* We settled on testing for the three sugars most discussed in scientific bean literature: raffinose, stachyose, and verbascose. (We learned in our tests that the pinto beans we were working with contain relatively little raffinose and verbascose but a lot of stachyose, so our analysis focused solely on that.)
*Here I feel the need for an aside. Many, many people will tell you that the key to reducing bean gas is to eat more beans. Eating more beans, they argue, works because it allows our digestive systems, and the microbiome in them, to acclimate to the beans. Over time, they say, the gassiness will go down. This makes no sense to me. If these oligosaccharides are food for bacteria in our gut, common sense would say that feeding that bacteria more food would, if anything, do the opposite by supporting their population growth while giving them plenty of raw material to digest. It wasn't within the scope of this project to test (and, I suspect, disprove) this theory, but count me as highly doubtful. If anything, I have to imagine that eating more beans more often just makes people more used to being gassy, and that, in turn, makes them notice it less. (Their significant others might have a very different take…)
What We Tested
Interview a hundred bean-eaters about how to minimize bean gas and you're likely to get two hundred suggestions. Use canned beans instead of dried! Rinse the canned beans! Soak dried beans! Discard the soaking water! Rinse your cooked beans! Throw some bay leaves in the pot! No, add a piece of kombu! It goes on and on. We decided to test several of these common recommendations, omitting some (e.g., turmeric, ginger) that fundamentally alter the flavor and/or appearance of the cooked beans in a way that may not be practical across a range of bean recipes. I wanted to test alkalinity—some people say baking soda makes for less gassy beans, but testing it properly would have added a slew of additional samples that we didn't have the bandwidth for (either way, baking soda can improve bean texture and reduce cooking times, so there's plenty of other reasons to add it; if it also reduces farts, I'd call that a bonus).
On top of all these dubious folk remedies, there's one that studies have shown actually works: Science has discovered an enzyme—alpha-galactosidase—that can break down the offending sugars into harmless bits. You can buy it at the pharmacy under the brand name Beano. According to package directions, one is meant to ingest a Beano tablet before eating beans. Problem solved. Not for me, though. I want a method to knock the wind out of beans that doesn't require me and my guests to pop a pill first. It's impractical. But, I wondered, could I move the alpha-galactosidase solution out of the medicine cabinet and into the pantry by using it not as a pill but as an ingredient when cooking the beans?
The Beano company flatly states that my idea won’t work: Cooking will denature the alpha-galactosidase enzyme before it does its job. I figured I could get around this blockade—that is pretty much what I do—by carefully adding the enzyme to the beans only when temperatures were low enough not to denature it.
In the end, we lab-tested 51 bean samples, measuring 17 different variables executed in triplicate. We opted for pinto beans because they’re widely available, taste great, and are anecdotally fairly farty. To see if this lab data made a difference to derrierès, the "Harvard Fart Squad" students then prepared a double-blind test of farty vs. non-farty beans (as determined by our first round of testing), handing out 45 cups of bean dip to their classmates. Here’s what we discovered.
How The Tests Were Run
I prepared all the necessary bean samples, blended them (since the samples need to be homogenous for the tests), vacuum bagged and froze them, and mailed them to Pia Sörensen and Kelly Chatman at Harvard’s "Small Molecule Mass Spectrometer" facility where they ran the quantitative tests with help from students Melody Cao and Fatema Abdulla. The description that follows is some real CSI-level stuff. The words are big, but the concepts aren’t complicated, so stick with me.
First, the oligosaccharides needed to be extracted from the bean puree by dissolving them in ultra-pure water. After that, centrifugation followed to remove the solids, and then the samples were run through a liquid chromatography/mass spectrometry unit. Liquid chromatography (LC for short) separates molecules based on how quickly they wash through a long column packed with solids resisting the flow. First, you run pure samples of the material—in this case, stachyose—through the machine to see how long it takes them to travel through the column. Next, you put in your unknown sample and wait that same amount of time for the stachyose to appear. Then, the mass spectrometer (Mass Spec or MS) takes over, vaporizing the stachyose into charged ions and using a voltage to accelerate those ions to high speed. The heavier those molecules are, the slower they travel, so by measuring their speed, you can determine their weight, thereby making sure you are only measuring stachyose. You are literally weighing individual molecules—pretty cool. The Mass Spec gave us a reading proportional to how much of the target molecule (in this case, stachyose) was present in a sample. After that, the crew at Harvard lyophilized (fancy word for freeze-dried) the beans to see how much water they contained.
Do Any Traditional (i.e., Non–Beano) CookingTechniques Reduce Bean-Induced Fartiness?
Categorically no. You can soak your beans for anywhere between 8 and 24 hours, you can throw away that soaking water or keep it, you can cook direct from dry, you can pressure cook, you can cook with bay leaves or kombu, you can even parboil soaked beans for a minute in boiling water and throw that water away. None of these strategies makes an appreciable difference in the fart levels of beans cooked from scratch. Strangely, the beans cooked with bay leaves had the highest fart potential of any of the samples tested—almost as though the bay conserved the farts rather than dispelling them. But we would need more tests to be sure.
These tests blew me away. At the very least, I expected that presoaking the beans and pitching the water would reduce fartiness. After all, the sugars are water soluble, so they should leach into the soaking water and get discarded. As it turns out, not so much. I still don’t understand this result.
Do Canned Beans Fart Less Than Those Cooked from Scratch?
Yes, marginally. We tested nine cans of Goya pinto beans from three different lots, and they consistently had about twenty percent less fartiness than the beans cooked from dry.
Note, though, that this is not a statement we can make with certainty because the canned beans were from Goya and the dried beans were from the Jack Rabbit company. It is entirely plausible that the Goya corporation has a source of pinto beans with less fartiness than the folks at Jack Rabbit.
Does Rinsing Canned Beans Reduce Gas?
A resounding yes. Gram per gram, rinsed beans are over twenty percent less farty than unrinsed, and the liquid you throw away is thirty percent fartier than the beans are themselves. The liquid in the bean can has many more farts per gram than the beans themselves do, but pitching this liquid comes at a high cost: flavor. Compared to beans with their cooking liquid, rinsed beans are tasteless. Here is why:
A 15-ounce can of pinto beans contains roughly 447 grams of beans and liquid. Drained and rinsed, you end up with 267 grams of beans. Getting rid of the liquid means you discard over a third of the can contents. Furthermore, the bean liquid contains about fifteen percent of the solids that the raw beans started with. But wait—it gets worse. The bean solids that are in the liquid are the ones that dissolve in water, which are also the solids with flavor. Think about making a meat broth: Simmering meat for a long time in water shifts the meat’s flavor into the liquid. Same with beans. Rinsing beans is like throwing away beef stock and eating only the flavorless, overcooked beef.
The amount of farts in your bean/water system is relatively constant. If you reduce the water, the farts per gram in your system go up. Conversely, if you dilute (as you would in a soup), the farts per gram decrease. Looking at the bean liquid this way may indicate why home-cooked (from dry) beans may seem fartier to you: When you cook beans from dry, you may end up with less liquid per bean than you have in canned beans. As we’ve seen, a can of beans is over one-third liquid, which you will typically toss or use. If you are cooking dry beans to use as a side dish, there's a chance you will reduce the liquid even more during cooking, and reducing the liquid increases the farts per gram.
Does Beano Work As a Gas-Reducing Ingredient in the Bean Pot? And Should You Use It?
Yes and no. And maybe. In our tests of pureed beans—think bean dip—Beano did a fantastic job. With whole beans, though, it appears that only the liquid gets effectively de-farted.
Pureed beans treated with Beano had almost two and a half times less fart potential than pureed canned beans and nearly three times less fartiness than beans cooked from scratch. A big win!
The technique is simple: Use one Beano 800 (800 is a measure of how much enzyme the Beano contains) per pound of cooked beans. So a 15-ounce can of beans takes one Beano (I use two, cause hey—better safe than farting). If you cook a pound of dried beans, you should end up with about four pounds of cooked beans and broth, so you’ll use four (or, in my case, eight, because, as I said, I'd rather not be farting). You pull apart the Beano gel capsules to harvest the powder inside. After you have cooked your beans from dry, wait for them to cool down to 104°F/40° and add the powder; the exact temperature isn’t essential as long as you don't add the enzyme when the water is hot enough to denature it (I don't actually know what that upper limit is, but 104°F is safe). If you are using canned beans, add the powder before you heat them. Puree the beans with the Beano in a blender and wait about an hour for the enzyme to do its thing. Then heat or eat.
Beano didn't do as well in our quantitative lab tests of whole beans. We added Beano to canned beans without draining the liquid and warmed the mix up to 104°F (40°C) for an hour to let the enzyme work. Then we heated the beans with their liquid to boiling to destroy the enzyme and separated the liquid from the beans for testing. The liquid was virtually fart-free, but the whole beans had only slightly less fartiness than canned beans that had been drained and rinsed. The upshot? When you use the Beano technique on whole (as opposed to pureed) beans, you are only treating the liquid. Essentially you are turning the unrinsed beans into rinsed beans while preserving the flavor. Worth doing if bean fartiness is an issue for you or your guests, but not enough to eliminate bean farts entirely.
The "Harvard Fart Squad" In-Vivo Data
With this quantitative lab data in hand, Ada Vazzana, Micaela Rosen, and Vivian Nha Nguyen, the students running the Fart Squad, went to work in the field. After all, it's one thing to see the oligosaccharide content of the beans as measured by lab equipment, but it's another to be able to say whether any observed differences in the lab have a measurable impact on real, live digestive systems.
Given that we saw the most significant effect on oligosaccharide reduction by treating pureed beans with Beano, we decided that was the variable to put to the test with our live subjects. The students prepared a bean dip using canned pintos, olive oil, lemon juice, and salt. Half the dip was treated with Beano following the protocol described above; half was not. They then handed the samples out randomly to 45 fellow students, who did not know whether they were receiving the Beano sample or the control. The students also received a questionnaire at class time, which was before lunch.
“The flatulence hit almost immediately after eating the dip.”
Volunteers rated themselves on how farty they believed themselves to be usually versus how farty they felt six hours after eating the bean dip. As of this writing, we have half of the surveys back, and even with this smaller-than-hoped-for sample, the Beano group reported, on average, slightly less gas after eating the beans than they usually have on any given day. In contrast, the non-Beano control group scored an average of one point fartier on a 10-point scale. One point doesn’t seem like much until you dig a bit deeper: Only one person in the Beano group felt two points gassier than usual. In contrast, one-third of the control group scored three or four points gassier, with one respondent saying, “The flatulence hit almost immediately after eating the dip.”
For bean dips and purees, Beano works. For everything else, fartgeddaboutit!
Addendum: The Cooking Procedures
- Cook from dry with no soaking (the control): Cook 100 grams of beans in 400 grams of water with 2.6 grams of salt for roughly two hours, adding water as necessary so that the final batch weighs 388 grams. Why did I choose these seemingly random numbers? Well, as it happens, that is approximately the same bean-to-water and bean-to-salt ratio I measured in my canned bean samples. I wanted the dry bean tests to be apples-to-apples with the canned beans. In cooking for myself, I would further reduce the final amount of liquid in the beans.
- Soak for 8 hours and then cook: Using the same ratios as the control above, soak the beans in the water and the salt for eight hours before cooking. The cooking, in this case, only took about an hour. I added salt to the soaking water because salt helps beans absorb more water while they are soaking. Rationale: some people believe that the beans—still alive when they are dry—will start to break down the complex sugars as they soak. Eight hours was the minimum amount of soaking time we thought might have an effect. As an aside, many bean people believe that presoaking makes for better beans. I don’t wish to get into this argument with you, but I’ll say this: Presoaking larger and tougher skinned beans makes beans cook much more quickly on the stove and helps keep the beans whole as they cook.
- Soak for 24 hours and then cook: Same as above, but soak the beans for 24 hours—the maximum time we thought was reasonable.
- Soak for 8 hours, throw away the soaking water, and cook: Same as number two, but I threw away the soaking water and added fresh water and salt before cooking. Rationale: since the offending complex sugars are water soluble, they should leach into the water and go down the drain. Many bean purists hate throwing away soaking water because it also discards flavor. I tasted all the water I threw out—it didn’t taste like much. Significant color leached into the water, though, and this would be a massive deal in dark beans like red kidney or black beans.
- Soak for 24 hours, throw away the soaking water, and cook: Same as above, but soak for 24 hours.
- Soak for 8 hours, throw away the soaking water and then blanch in boiling water for 60 seconds; discard blanch water, and cook: Same as number four, but add a blanching step before cooking. Rationale: Canned beans are most often blanched in near-boiling water before being canned to get rid of any air inside the bean that would cause problems during canning. We theorized this quick blanch would remove toots. Interestingly, these beans tasted blander than the other beans cooked from scratch (I tasted every sample), with a flavor closer to the canned beans.
- Soak for 8 hours with bay leaves and then cook: Same as number two, but two bay leaves were added to the soaking water and allowed to remain during cooking. Rationale: Some people believe that bay leaves mellow beans. I told Daniel that I regularly use a ton of bay leaves because the ones I buy are low quality and lack flavor. He told me to get decent ones and use an average amount instead. The ones I purchased were from La Boite Epice and were, in fact, revelatory. I was shocked at how much better they were than my supermarket dross. Two leaves in the standard recipe made the entire batch redolent of bay. I am now firmly sold on only using good-quality bay leaves.
- Soak for 8 hours with kombu and then cook: Same as above, but using kombu instead of bay. Kombu is the umami-rich seaweed used to make the Japanese stock dashi. Many people believe kombu helps lessen the windy burden of the bean. The kombu I used was the Wellpac brand from Korea, and I used a 120x115 mm square (6.55 grams) in each batch. Surprisingly, the beans did not taste that much of kombu, although they were perhaps a bit richer.
- Pressure cook without soaking: Rationale: the high heat in pressure cooking (259°F as opposed to 212°F) might break the sugars down into harmless bits. All tests were cooked in a Kuhn Rikon pressure cooker at second ring (15psi) for 45 minutes and allowed to cool naturally before I opened the cooker.
- Soak for 8 hours and then cook, cool, and add Beano: Last, my piece de resistance, my secret weapon in the bean wars, and the main reason I wanted to run this study: Beano. Cook as in number two. After cooking, wait for the beans to cool to 104°F and stir in two capsules of Beano powder.
Cover illustration by Michelle Kondrich