The Quest for "Forbidden Nut Flavor"

On December 11, 2018, the United States Patent and Trademark Office granted Patent Number 10,150,935 for a seemingly wild new culinary technique. The patent outlines a procedure for running non-hydrogenated almond butter through a proprietary steam distillation process to yield a concentrated extract with the full, rich flavor of almond, but none of an almond's allergenic properties. This new flavoring agent would offer people with almond allergies the ability to "enjoy the desired and otherwise forbidden nut flavor," the patent claims, not just imitations of it.

Allergy-Safe Real Almond Flavor?

Allergic reactions to so-called tree nuts are among the most dangerous allergic reactions, accounting for up to 40% of food-related anaphylactic episodes. And almonds are among the most allergenic of all tree nuts. So the idea of using real almonds to create a supposedly allergy-safe almond flavoring may seem absurdly risky.

But the process detailed in this patent isn't some lark daydreamed into existence by a reckless basement inventor. Several allergy and food science experts I've spoken to note that we already know how to turn allergenic raw ingredients into non-allergenic products, albeit not flavorful ones. This patent's main author is Terry T. Kirihara, who was a chemist and researcher at General Mills at the time he came up with the idea, and instead of filing the patent as an individual, he assigned the intellectual property to Big G.

The patent repeatedly notes that the extract would be ideal for use in "a widely popular breakfast or Ready-to-Eat puffed cereal…that comprises a cereal base in the form of puffed rings fabricated from a cooked oat dough…The cooked cereal dough and/or sugar coating importantly includes a characterizing nut flavor provided by a honey roasted almond nut butter ingredient."

All of this suggests that the process was developed with a specific product in mind: Honey Nut Cheerios. Why, then, haven't we heard more from General Mills about this mind-boggling technique, which could potentially unlock new worlds of flavor and safety for people with certain allergies?

To understand the apparent lack of fanfare around the ideas outlined in this patent, you need to understand the cutthroat world of breakfast cereal economics and marketing—and the confusing frontiers of allergy science and food safety regulations.

I reached out to General Mills to see if the now-retired Kirihara or anyone else at the company would be able to tell me a little more about this patent. They declined to comment without specifying any reason why. When I tried to follow up a while later, they did not reply.

Honey Nut Cheerios, Hold the Nuts

Public records show that Kirihara, who'd reportedly worked on Cheerios projects since at least the 1990s, apparently filed the patent in late 2006—around the time that General Mills reportedly stopped using real almonds in Honey Nut Cheerios, silently switching over to "natural almond flavor."

Food regulations state that companies don't have to tell consumers the specific ingredients used to create any "natural flavor." Many manufacturers use this vague term in ingredient lists to hide their use of usually cheap taste-alike replacements for expensive or scarce ingredients.

Producers frequently replace almonds with ground-up apricot or peach pits, which they list as "natural almond flavors." Although we call them nuts, almonds (and walnuts, pecans, and whole nutmeg) are actually the pits of a fruit closely related to these and several other "stone fruits," like cherries, nectarines, and plums. Apricot and peach pits specifically share many of almonds' key flavor compounds and are otherwise usually considered waste products.

Reporters long assumed General Mills made this switch to cut costs, and that they've been using one of these pits in Honey Nut Cheerios ever since. The patent suggests that they may be right. "While adding desirable taste and nutrition," the patent notes, "almond nut butter is an expensive ingredient typically costing up to 5-20 times the cost of the oat flour and other ingredients" used to make the popular cereal it references. "The cost of this one minor (in terms of weight) ingredient can be in the range of the total cost for all the other cereal ingredients combined."

"With ready-to-eat cereals, cost savings are important," says Joshua Berning, a food marketing expert at Colorado State University who has studied breakfast cereals. After all, their makers are competing in a crowded market for often price-sensitive buyers. It's no surprise that many big brands have abandoned expensive real ingredients for cheap "natural flavors" in recent years.

But even though this sort of silent switch-out was and is a common tactic, Kirihara's patent suggests he and others still worried about the potential risks of transitioning to a taste-alike. "Some consumers when reading the ingredient listing for food products find such substitution to be 'dishonest' or 'misleading,'" the patent notes. "Also, some more discriminating consumers are able, or believe themselves able, to distinguish the taste of such substitutes from real almond flavors."

(True to this prediction, when news of Cheerios' sneaky switch-up hit social media and food blogs in 2014, then again in 2016, many writers and readers expressed a sense of betrayal.)

The patent itself throws serious shade at "'almond' flavors derived from apricot pit extracts," which it identifies as a major, likely alternative for the cereal in question. It opines that these "phony flavors…are merely reminiscent or merely mimic the flavors."

Making the switch to "natural almond flavor" didn't even lead General Mills to take its allergy warning off of Honey Nut Cheerios, which could have opened the cereal up to the small but lucrative allergic consumer market: While 1% of Americans at most have a tree nut allergy, and only about 10 to 15% of them are allergic to almonds, allergy rates are on the rise, and whole families often go allergy-free to support one members' dietary needs. Boxes still read, "CONTAINS ALMOND INGREDIENTS."

No one I've spoken to is sure why that is. It could be that Cheerios' natural flavoring still has a pinch of almonds in it, as a few other companies' additives reportedly do. It could be that the company doesn't want to publicize its abandonment of almonds, for fears of brand damage. It could be that they still produce Honey Nut Cheerios in factories that handle almonds for other products. Or it could be that they're acting on scattered reports of people with tree nut allergies having bad reactions to apricot and peach pits. There's not much research on the risk of reactions to these ingredients; an FDA representative told me they haven't issued any specific guidance on their potential allergenicity. But companies often play it safe when labeling their products to avoid possible public relations nightmares and the risk of having to do costly recalls.

Kirihara's proposed process seemingly attempts to balance cost and authenticity concerns. The patent points out that distilling pure flavor into an extract could potentially be more aromatically cost-efficient than using raw nut butter. It would also leave behind de-flavored butter the company could use in other products, possibly as a bland protein base, or sell to others "at substantially equivalent prices as conventional nut butters." The extract, the patent argues, would also create "a natural full-bodied flavor rather than the 'thin' flavor provided by almond flavor substitutes from apricot pit extracts," potentially giving them a leg up over their competition among some consumers.

The patent adds that this extract would also finally allow the o-shaped cereal in question to drop its allergy warning labels, and all of the costly anti-contamination measures involved in making a food item that contains an allergen, drawing in new consumers and further cutting costs. "If this process reduces cross-contamination [mitigation] costs, that's an advantage," Berning says.

As "new flavors, ingredients, and approaches can provide a slight advantage over competition," Berning explains, cereal makers constantly spitball innovations. Kirihara alone has filed numerous patents that he transferred on to Big G, for everything from high-protein "puffed food products" to "laminated multi-layered cereal products." 

Taking Out Allergens

Kirihara's patent for a means of creating a non-allergenic almond flavoring doesn't read like a wild swing of an idea, though, in part because it's actually fairly easy to make non-allergenic foodstuffs out of allergenic raw ingredients. You've probably consumed some already.

Allergies are immune system (over)reactions to specific proteins within foods. If you destroy those proteins or sieve them out, then you'll put an end to any allergy risk they pose. You don't even need to tackle all the allergenic proteins in an item; below a given parts-per-billion concentration, even hyper-sensitive immune systems won't react to them.

Neutralizing specific proteins is tricky, as a Spanish research group noted in a recent review of techniques. The effects of heat, pressure, radiation, and any other intervention you can think of "depend on the intrinsic characteristics of the protein and the type and duration" of the technique, they write. A process that damages, destroys, or culls one allergenic protein might do nothing to another—or even warp some proteins in ways that make them more allergenic.

But while we're not great at surgical strikes against allergens, we have mastered scorched-earth tactics. An FDA representative said that US food safety laws specifically exempt highly refined oils made of common allergenic foods from allergy labeling because the processes used to extract those oils seem to remove or destroy enough of their allergy-causing proteins to make them reliably safe for everyone. (However, the representative added a worrying caveat: the agency doesn't actually have a hard-and-fast definition for what constitutes a highly-refined oil, just a general list of processes that can highly refine things.) It's hard to find these oils in grocery stores, but highly-refined peanut oils are especially common in industrial cooking, either on their own or within vegetable oil blends, thanks to their low cost and high smoke point. Chances are good that if you've eaten fried food at a restaurant you've had some of that oil.

The FDA doesn't deal with allergen risks in alcoholic beverages. But the American Academy of Allergy, Asthma & Immunology notes that spirits flavored with nut products before distillation, like the often almond-laced amaretto, are almost certainly (but not reliably or definitively!) non-allergenic as well, as the distillation process ultimately destroys or removes most of their allergens.

No one I spoke to for this story could say, given the amount of information in Kirihara's patent, how effective his process would be relative to others in destroying or separating out allergenic proteins. However, its reliance on heat-based distillation certainly shows broad potential.

"I would worry about the process leaving something behind that is still a problem," cautions Scott Sicherer, a doctor in New York's Mount Sinai medical system who specializes in food allergies.

But Stephen Taylor, a food allergy researcher at the University of Nebraska, says that if the patent's claims about its process's ability to separate out allergenic compounds bear out, "then this ingredient would not pose any risk to almond-allergic individuals."

Despite the granting of the patent and the apparent commercial logic behind it, there's no clear sign that Kirihara's process is being used. This could just mean that General Mills is still slowly working on getting it ready for prime time. It's also possible that Big G has already started to use this approach—just without any fanfare or attempt to capitalize on any claims to hypoallergenic status. But it seems more likely that the idea ran aground on market and regulatory uncertainties.

Flavor vs. Risk

Susie Bautista, a flavor chemist and food safety expert, points out that even though scientists are getting better at identifying and isolating the compounds that give foods their distinctive flavors, it's still hard to "imagine that it's possible to isolate all the compounds associated with aroma and flavor in a food item." Each compound requires specific conditions for ideal extraction, and every process runs the risk of some degree of adulteration, which could have subtle effects on flavor.

So even if Kirihara's proposed approach broadly works, there's no guarantee that it would lead to an exact taste translation. Bautista suggests that General Mills would have to carry out costly consumer taste tests to figure out how close they got, both in general and relative to the "natural almond flavor" they're using. Then, they'd need to see how much value consumers placed on any taste gains the process yielded, relative to the costs associated with the new process.

The FDA also says that it has not set any clear benchmarks or thresholds for declaring a new ingredient hypoallergenic. If General Mills would want to follow the patent's advice and make that claim, it would need to seek an exemption from allergen labeling laws. Taylor notes that the process would ultimately involve intensive tests to measure the protein levels in Kirihara's extract, then independent reviews of those studies and their findings—all of which would be expensive. “It's not an easy process,” he stresses.

Even if General Mills proved this almond extract has no discernable levels of known almond allergens, they still might think twice about dropping their allergy warning labels in a search for new customers. While researchers have identified six distinct allergenic proteins in almonds, they're not certain that they've identified every potentially allergenic protein in these nuts or most others. There's always a risk that down the road someone proves one of the components left in the extract is in fact a minor allergen that affects small groups—which may not lead to many health issues but would still create regulatory and public relations headaches for Big G. 

And even if the extract is proven to be completely hypoallergenic, that doesn't mean almond-allergic consumers will flock to it. Sicherer points out that even people who outgrow their almond allergies (as about 9% eventually do) often still avoid them as a matter of habit, taste, or extreme caution. "Some almond-allergic consumers develop a taste aversion to any almond flavors because they can trigger an anxiety response," says Taylor.

On top of all of this, Berning's not sure, based on his experiences with cereal consumers, how many of them actually care about finding real nut flavors.

"Firms have to expect to recoup their investment in a new process, and ultimately earn a return on it, too," Berning explains. That kind of return is far from certain, given all of these variables.

Steve Leusner, a cereal consultant who used to work at General Mills, went so far as to argue that Kirihara's idea brings so little surefire gain to the table that "it makes no sense at all."

These are classic pitfalls of commercial research and development. For lack of a clear and strong market incentive, dozens of fascinating food innovations end up consigned to the trash heap of history every year. Similar uncertainties have ended several other efforts to develop nut allergy-safe nut products—many more ambitious than Kirihara's extract.

Most notably, in the 2010s researchers at a food tech firm conducted experiments with a new process using an enzyme bath to break down the allergenic compounds in whole, skinless peanuts. Skin prick tests on human subjects were promising, but they reportedly failed to eliminate all allergic reactions—and reports on their product never really mentioned taste. Other research teams have spent decades using gene editing techniques to turn off the genetic signals that lead to the production of allergenic proteins. However, at best, their approaches have eliminated some but not all of the allergenic proteins in peanuts. 

As a result, despite all the work poured into these projects to make allergy-safe nut products overall, "there are no commercially successful hypoallergenic nut products on the market," Taylor points out. "Unless you want to count highly-refined peanut oil."

Fortunately, even if individual projects hit dead ends, food companies are still more than eager to keep pumping money into blue sky projects, in the hopes that eventually they'll crack the nut and land upon some big innovation that'll help them stand out in the packed aisles of your local grocery store. So, even if Kirihara's patent isn't destined to move into active production, there's still a chance that one day he or one of his colleagues will eventually bring that "forbidden nut flavor" to the masses.