The Science · The Library  ·  By Sara Kirkwood, Founder of Improve Equine  ·  June 2026

“Natural” and “proven” are not the same word, and neither are “a plant” and “a drug.” Before we ever get to which herbs do what in a horse, it’s worth slowing down on what these words actually mean — because the space between them is where most of the confusion in the supplement aisle lives.

Zoopharmacognosy. Say it five times fast.

It’s the word people reach for when they talk about animals and plant medicine, and it sounds impressive enough to end an argument. But it’s worth knowing what it actually means, because the word is doing a lot of quiet work in the herbal-supplement world — some of it earned, some of it borrowed.

Zoopharmacognosy is a real, peer-reviewed field of study. The name was coined in the early 1990s by chemical ecologist Eloy Rodriguez and primatologist Richard Wrangham, stitched together from Greek roots — zoo (animal), pharma (drug), gnosy (knowing) — to mean, roughly, “animals knowing medicine.” And it studies exactly that: wild animals selecting and using plants, soils, even insects to treat themselves.

The evidence is genuinely remarkable. Wild chimpanzees swallow certain rough, bristly leaves whole, without chewing — and the stiff hairs on those leaves physically scour parasitic worms from the gut wall, sweeping them out within hours. The behavior was documented across multiple populations and tied directly to the control of nematode infections (Huffman et al., 1996). Sick chimps have also been seen chewing the intensely bitter pith of Vernonia amygdalina — a plant local healers use for the same parasites — and recovering over the following days (Huffman & Seifu, 1989). Honeybees gather antimicrobial plant resin into the hive, and colonies with that resin envelope show measurably lower expression of costly immune genes — they get sick less, so they spend less defending themselves (Simone, Evans & Spivak, 2009).

But notice how narrow that is. Every one of those examples is a wild animal, while sick, using a specific substance against a specific ailment — and the researchers who study this hold it to strict criteria precisely so it doesn’t get confused with ordinary eating. It is careful, bounded science. When you see “nature knows” or “horses choose their own medicine” stretched into a sales pitch, that’s the impressive word being asked to carry weight it was never built to hold.

The word that actually matters: pharmacognosy

Drop the “zoo” and you get the field that genuinely underwrites this entire conversation: pharmacognosy — the science of drugs derived from natural sources. It’s old and serious, named back in 1811, and it’s the study of what’s actually in a plant, animal, or fungus that does something in a body, and how to identify, isolate, and standardize it.

I notice a tendency to split the world into “natural” over here and “drugs” over there, as if a medication is something purely man-made in a lab and therefore a little suspect, while an herb is innocent. That split is mostly a myth. An enormous share of modern medicine didn’t escape from nature — it came out of it. Aspirin traces to willow bark. Morphine comes from the poppy. Artemisinin, the frontline malaria drug, comes from sweet wormwood — and won its discoverer a Nobel Prize in 2015. The heart medication digoxin comes from foxglove. This isn’t a handful of charming exceptions: in the most careful accounting we have, roughly two-thirds of the small-molecule drugs approved between 1981 and 2019 were natural products, derived from them, or directly inspired by them (Newman & Cragg, 2020).

So “it’s just a plant” and “it’s a drug” are not opposites. Very often they’re the same compound at different points in its story — the plant is where we found the active ingredient, and the drug is what happened once we isolated it, concentrated it, and delivered it on purpose. The thing that turns a plant compound into a medicine isn’t magic and it isn’t man-made wickedness — it’s dose. Isolated, concentrated, standardized, and prescribed at a level chosen to produce an effect.

Nature is not gentle

There’s an assumption tucked underneath the word “natural” that I want to pull out into the light, because it quietly drives a lot of decisions: the feeling that natural means benevolent. Gentle. On our side. Nature as a nurturing force that only ever wants to heal.

This is a false narrative.

Nature is not gentle. Nature is a constant, savage negotiation between things trying to grow and things trying to eat them. And a plant is in a particularly tight spot, because a plant cannot run. It can’t flee a grazing mouth, it can’t swat an insect, it can’t hide. So over millions of years, plants evolved the only defense available to something rooted in place: chemistry. They became expert poisoners.

The compounds we prize in medicinal plants — the alkaloids, the glycosides, the bitter and the pungent and the aromatic — are, overwhelmingly, the plant’s chemical weapons: substances it manufactures specifically to sicken, deter, or kill the animals and microbes that would otherwise consume it (Wink, 2003). Caffeine is a natural insecticide. Nicotine is a plant toxin. The active compound in foxglove that steadies a failing human heart will also stop one. These molecules are potent in animal bodies precisely because they were engineered by evolution to disrupt animal biology. That is the whole reason they “do something.”

Which brings us right back to dose, and to the oldest rule in pharmacology. Five hundred years ago, the physician Paracelsus put it in one line: the dose makes the poison. A lot of these natural ingredients are toxic at some dose, therapeutic in a narrow band, and harmless below it. The line between medicine and poison is not a wall between two different substances. It is a dial on the same substance. That’s not a reason for fear — it’s the reason dose is the only serious question. “Natural” tells you nothing about where you are on that dial. Only the amount does.

And there’s one more wrinkle worth naming, because it matters for how we should think about whole herbs. When a pharmacologist isolates a single active compound, they know precisely what they’re dosing. When you hand a horse a whole herb, you’re handing over the isolated compound and everything else the plant built alongside it — fiber, other chemistry, buffers, compounds we haven’t even characterized. Sometimes that whole-plant matrix softens the edges of the active compound. Sometimes it adds effects of its own. It’s a genuinely interesting distinction, and it’s a big enough subject that it deserves its own article rather than a paragraph here. For now it’s enough to say: a whole herb is not a clean single dose of anything, and that cuts both ways.

And then there’s the hopeful version

There’s a third thing in this space, and I want to talk about it gently, because the people drawn to it have their hearts in exactly the right place. It often goes by “applied zoopharmacognosy” — offering a horse a tray of herbs or essential oils and letting the horse “self-select” what it supposedly needs, on the theory that the animal’s instinct is reaching for medicine.

I understand the pull of that completely. It’s a beautiful idea: that your horse knows its own body, that instinct will guide it to exactly what will help. And here’s the honest, interesting truth — there is real science on animals selecting beneficial plants. It’s just earthier and less woo woo than the marketing suggests. In careful studies, parasitized sheep will, under some conditions, increase their intake of tannin-rich forage that helps suppress their worm burden (Villalba, Provenza & Shaw, 2006). But when researchers looked closely at how, the mechanism wasn’t a mystical inner knowing — it was learning. The animal samples something, its body registers whether it felt better afterward, and it forms a preference through that feedback over time. And even then the results are mixed: some well-designed studies find the effect, others don’t find it at all. It’s a real, fragile, learned phenomenon — and nearly all of it has been shown in ruminants like sheep and goats, whose digestive lives are very different from a horse’s.

So where does that leave the horse’s bucket? Honestly: as a beautiful idea that hasn’t been vetted. Not debunked — vetted. Nobody has run those studies in horses, which means it sits under a big question mark. It might be doing something real. It might be nothing more than a horse picking what tastes good. We don’t know, because the work to find out hasn’t been done.

And that question mark is exactly where your own reason for doing it starts to matter. So ask yourself honestly which one you’re standing on:

If you’re doing it because you believe it’s effective — because you think the horse is treating a real problem and you’re relying on that — then you owe it the science. Is there evidence in horses? Is it the right dose? Is the compound even absorbed? When your horse’s health is riding on something working, a good feeling isn’t enough; that’s precisely the moment to ask for verification, because being wrong has a cost.

But if you’re doing it because it’s a kind, pleasant thing your horse plainly enjoys — no treatment claim attached, just a nice moment offered to an animal you love — then that is already reason enough, all by itself. It doesn’t need to be dressed up as medicine. It doesn’t have to move mountains. A thing can be healthy, good for the horse, and something the horse genuinely likes, and that combination is a perfectly good reason to do it.

The whole trick is just being honest with yourself about which of those two you’re actually doing — because the moment you’re relying on it to treat something, the bar quietly moves from “the horse enjoys it” to “show me the evidence and the dose.” That’s a much more comfortable, and a much more truthful, place to stand.

Where this leaves us

Three words that get tangled together, pulled apart: zoopharmacognosy, the careful science of wild animals treating themselves; pharmacognosy, the rigorous field that turns plant compounds into measured medicine; and the hopeful, largely untested practice that borrows the first one’s name and the second one’s credibility without having earned either in a horse.

Underneath all three is the same quiet truth. Nature is not a gentle pharmacy with your horse’s name on the prescriptions. It’s a brilliant, indifferent chemist, and the compounds it makes are powerful exactly because they were built to act on animal bodies. What decides whether one of them helps your horse, does nothing, or does harm is never the word “natural.” It’s the dose, the evidence, and the rigor to answer the question you’re really asking.

Frequently asked questions

References

• Huffman, M.A., & Seifu, M. (1989). Observations on the illness and consumption of a possibly medicinal plant Vernonia amygdalina (Del.), by a wild chimpanzee in the Mahale Mountains National Park, Tanzania. Primates, 30: 51–63.
• Huffman, M.A., Page, J.E., Sukhdeo, M.V.K., Gotoh, S., Kalunde, M.S., Chandrasiri, T., & Towers, G.H.N. (1996). Leaf-swallowing by chimpanzees: a behavioral adaptation for the control of strongyle nematode infections. International Journal of Primatology, 17: 475–503.
• Simone, M., Evans, J.D., & Spivak, M. (2009). Resin collection and social immunity in honey bees. Evolution, 63(11): 3016–3022.
• Newman, D.J., & Cragg, G.M. (2020). Natural products as sources of new drugs over the nearly four decades from 01/1981 to 09/2019. Journal of Natural Products, 83(3): 770–803.
• Wink, M. (2003). Evolution of secondary metabolites from an ecological and molecular phylogenetic perspective. Phytochemistry, 64(1): 3–19.
• Villalba, J.J., Provenza, F.D., & Shaw, R. (2006). Sheep self-medicate when challenged with illness-inducing foods. Animal Behaviour, 71(5): 1131–1139.
• Villalba, J.J., Provenza, F.D., Hall, J.O., & Lisonbee, L.D. (2010). Selection of tannins by sheep in response to gastrointestinal nematode infection. Journal of Animal Science, 88(6): 2189–2198.

The Science · The Library  ·  By Sara Kirkwood, Founder of Improve Equine  ·  June 2026

Supplements live in a buyer-beware market. A claim on the label and a result in the bucket may be two very different things, and the distance between them is where I invest my attention.

Start with the study, not the label

A claim is only as good as the research underneath it. The standard I want is peer-reviewed: a hypothesis that was tested, with results that were reproduced under stated conditions. The gold standard is a randomized, double-blind, placebo-controlled trial on horses. If a company markets an ingredient but can’t point to a study on that ingredient, in the form they’re actually using, on the species they are treating, they’re selling a hypothesis dressed as a solution. That doesn’t make the ingredient useless. It means the evidence isn’t there yet, and I get to decide what I do with that.

This is also where a proprietary blend gets in the way. I understand wanting to protect a formula — the recipe is intellectual property, and nobody wants their work copied. But the effect of “proprietary” is that the per-ingredient amounts are withheld, and those amounts are exactly what I need to run every test in this article: the dose, the form, whether the active clears a level that does anything.

Did they put in the dose the study used?

Here’s where most products quietly fall apart. A study that proves an effect proves it at a specific dose. If the research shows that 5 grams a day produced a response, and the scoop in front of me delivers 2.5, the label still gets to list the ingredient, but the protocol that produced the result is broken. I might get some benefit at half the dose, but I am no longer reproducing the results I was sold on.

Will the horse actually absorb it?

An ingredient on the label is not the same as an ingredient in the bloodstream. The same compound, in a different chemical form, can reach the body at wildly different rates, and quercetin is the clearest example I know of.

In its plain, free form, quercetin is poorly absorbed. In dogs, its absolute oral bioavailability came in at roughly 4 percent (Reinboth et al., 2010, British Journal of Nutrition) — the overwhelming majority never reached circulation. Attach it to a sugar and the picture changes. In rats, plain quercetin landed at about 2 percent, the glucoside form isoquercitrin reached 12 percent, and an enzymatically modified version reached 35 percent (Makino et al., 2009, Biological and Pharmaceutical Bulletin). And in people, a lecithin-based formulation reached plasma levels up to twenty times higher than the same dose of unformulated quercetin (Riva et al., 2019, European Journal of Drug Metabolism and Pharmacokinetics).

So “500 mg of quercetin” on two different labels can mean two completely different things in the horse, without either label saying a single false word. The form is the part you have to read for. (And worth noticing: every number I just gave you is from a dog, a rat, or a person. Hold that thought — it comes back at the end.)

Does it survive the bag, and then the stomach?

A compound only counts if it’s still active when it reaches the place it gets absorbed, and there are two stretches of road where it can erode along the way.

The first is the shelf. Plenty of raw ingredients break down on contact with light, oxygen, or moisture. An antioxidant that has been exposed to air in the tub in your feed room is degraded before it ever hits the bucket. Processing is part of this too. Pelletizing a supplement is convenient, and it keeps a horse from sifting out the powder they don’t like, but it applies heat and steam and then pressure, and heat-sensitive actives can degrade in the process. In pelleted swine diets, recovery of vitamin K fell to about 75 percent under high-temperature, high-compression pelleting, compared with roughly 84 percent under gentler conditions (Wang et al., 2021, Animals). What went into the mixer is not always what’s in the bag.

The second stretch is the gut itself. The equine stomach is an acidic environment, and some compounds and most unprotected probiotics are degraded there before they reach the small intestine or hindgut where absorption actually happens. Feed lactobacilli tested against simulated gastric and bile conditions survived poorly unless they were buffered by a protective carrier (De Angelis et al., 2006, Research in Microbiology). “It’s in there” and “it arrived intact” are two different claims.

When more is just more

Absorption has a ceiling. Every nutrient has a saturation point, where the body’s transporters are full and additional ingredient simply passes through. Past that point, more of the ingredient isn’t more effect. It’s more expense, leaving the horse in urine or manure.

This is why a label boasting two or three times the amount of an ingredient isn’t automatically stronger. If the horse saturates well below that number, the extra is doing nothing but making the tub feel generous. And a cheap ingredient is easy to overload — if it costs the company almost nothing, they can put two and a half times the useful amount in there all day long and let the big number do the selling. A high number on the label may be a marketing decision and not a biological one.

The right tool for the outcome you want

Two products can share a shelf label and be built for completely different jobs. Joint supplements are the example I keep coming back to with my own horses. Supporting a young, sound joint to keep it healthy as long as possible is one goal. Keeping a geriatric horse with existing damage comfortable, slowing further degeneration, and calming inflammation is a different goal. The ingredients index on different mechanisms — some on cushioning, some on inflammation, some on oxidative stress — and the right one depends entirely on the outcome you’re actually after. Buying a preventive cartilage-support product for an arthritic senior is using the wrong tool, even if it’s a good formula.

The same is true even when two ingredients aim at the identical target. Quercetin and cromolyn both calm mast cells, the cells that drive allergic responses, but they don’t work the same way: in cultured human mast cells, quercetin worked prophylactically, building protection before exposure, while cromolyn had to be present at the moment of the trigger or it rapidly lost effect (Weng et al., 2012, PLOS ONE). Same destination, different route. Which one fits depends on whether you’re preventing a problem or reacting to one.

Who was actually in the study?

When a company points to a study, the next question is simple: proven on whom? A real finding in the wrong population can still be the wrong answer for your horse. Joint research, for instance, often runs on young, fit racehorses — animals with acute, exercise-driven inflammation, excellent blood supply, and a huge capacity to repair themselves. A result in that group doesn’t automatically carry over to a twenty-two-year-old with chronic, structural change. Different body, different biology, potentially different outcome.

And here’s the sharpest version of it, the one I promised to come back to. A great deal of the most-cited supplement science isn’t equine at all. The quercetin absorption numbers above came from dogs, rats, and people. The allergy research behind several popular ingredients — spirulina included, where the supporting trial was conducted in people with allergic rhinitis (Cingi et al., 2008, European Archives of Oto-Rhino-Laryngology) — lives almost entirely outside the horse. The underlying biology may well translate. But “may” is the honest word, and a study population that looks nothing like your horse is a reason to ask more questions, not fewer.

What I’m actually buying

None of this means supplements don’t work, or that the industry is out to get you. It means the label is written to balance information and marketing, and the answers that decide whether a product works aren’t always volunteered. Citation of a real study. Verified inclusion of the therapeutic dose in every serving. Presented in a form the horse can absorb. Packaged to survive the journey to the gut. Not over the saturation ceiling so I’m paying for what I don’t need. The mechanism matched to the outcome I want. Studied in something that resembles my horse. When you have all of those checked off, you’re buying results you can reasonably expect to reproduce.

References

• Reinboth, M., Wolffram, S., Abraham, G., Ungemach, F.R., & Cermak, R. (2010). Oral bioavailability of quercetin from different quercetin glycosides in dogs. British Journal of Nutrition, 104(2): 198–203.
• Makino, T., Shimizu, R., Kanemaru, M., et al. (2009). Enzymatically modified isoquercitrin is absorbed more easily than other quercetin glycosides or aglycone after oral administration in rats. Biological and Pharmaceutical Bulletin, 32(12): 2034–2040.
• Riva, A., Ronchi, M., Petrangolini, G., et al. (2019). Improved Oral Absorption of Quercetin from Quercetin Phytosome, a New Delivery System Based on Food Grade Lecithin. European Journal of Drug Metabolism and Pharmacokinetics, 44(2): 169–177.
• Wang, H., Yang, P., Li, L., Zhang, N., & Ma, Y. (2021). Effects of Sources or Formulations of Vitamin K on Its Stability during Extrusion or Pelleting in Swine Feed. Animals, 11(3): 633.
• De Angelis, M., Siragusa, S., Berloco, M., et al. (2006). Selection of potential probiotic lactobacilli from pig feces to be used as additives in pelleted feeding. Research in Microbiology, 157(8): 792–801.
• Weng, Z., Zhang, B., Asadi, S., et al. (2012). Quercetin is more effective than cromolyn in blocking human mast cell cytokine release and inhibits contact dermatitis and photosensitivity in humans. PLOS ONE, 7(3): e33805.
• Cingi, C., Conk-Dalay, M., Cakli, H., & Bal, C. (2008). The effects of spirulina on allergic rhinitis. European Archives of Oto-Rhino-Laryngology, 265(10): 1219–1223.

A few definitions, because this can all be confusing

Active ingredient. The part of a listed ingredient that actually does the work — not the total weight printed on the label. The two are not necessarily the same number.

Concentration. What percent of the raw ingredient is the active. A DHA-from-algae meal might be only a fraction DHA by weight; the rest is everything else in the meal.

Bioavailability. Of the active that’s actually present, the fraction that crosses into the bloodstream. This is a second discount, applied on top of concentration.

Bioactive form. The same molecule can exist in different chemical forms — bound to a sugar, formulated with lecithin, left plain — and the form changes how much is absorbed. “Quercetin” on two labels can behave completely differently for this reason.

Dose. The amount actually delivered per serving. The number that matters is the dose of active, not the dose of listed ingredient.

Therapeutic dose. The amount of active, reaching circulation, that a study showed produced the effect. Below it, you may have something in the bucket, but not the thing the study measured.

Sub-therapeutic dose. An amount that’s present on the label but sits below the level that does anything measurable. Common, and legal, and the reason “it’s in there” isn’t the same as “it works.”

Efficacy. Evidence that the ingredient produces the intended effect under controlled conditions. Separate from whether a given product contains enough of it, in the right form, to reproduce that effect.

Saturation point. The ceiling past which the body can’t use any more. Above it, extra ingredient is expense, not effect.

The Science · The Library  ·  By Sara Kirkwood, Founder of Improve Equine  ·  June 2026

A horse loses minerals in a very specific pattern when it sweats and most of what we reach for to replace them may be solving a problem the horse didn’t have — or worse, creating one. So let’s start from what’s actually leaving the body, and work forward from there.

This is filed under “The Science” because that’s where this ends up but I want to lay my bias on the table before we start, because it’s the thing I’ve asked the research to either confirm or correct. My instinct is that electrolyte supplements get reached for far more often than the horse in front of us actually warrants and that a lot of what gets sold as electrolyte replacement is really a drinking strategy in disguise — tip the chemistry, make the horse thirsty, hope it drinks. If true, a good portion of what we pour into a bucket ends up as expensive urine.

I asked the research to push back on me, and it did… but the core of it held up better than I expected so let’s build the whole thing from the ground. This is a peek into how I educate myself so you understand that questioning my own assumptions is where I start — because the goal isn’t to be right. It’s to get it right.

First, why any of this matters at all

Before we talk about what’s leaving the body, it’s worth a minute on why it matters that anything’s leaving at all. Salt and the trace minerals aren’t only nutrition sitting in the feed pan. Dissolved in the blood and in the fluid around every cell, they are the electrical system of the animal. Sodium and potassium moving across cell membranes are how a nerve fires and how a muscle contracts and then lets go — including the largest, most important muscle on the roster, the heart. Calcium and magnesium help time those signals. Chloride holds fluid balance and the body’s acid-base balance steady. When someone says a horse “needs electrolytes,” this is the actual stakes underneath the word.

The body runs all of that inside a remarkably narrow window, and it defends that window hard. Equine blood sits at a pH right around 7.4 and the body will spend real resources to keep it there. The horse is not passively losing and topping up minerals like a leaky bucket. It is constantly hunting for balance. Sweat carries minerals and water out. Thirst pulls water back in. The gut draws minerals from forage and feed. The kidneys quietly trim whatever’s in excess. Drink, sweat, eat, filter, repeat — that entire loop is one long balancing act, a body forever seeking its own equilibrium and never quite sitting still at it.

This is exactly why supplementing is a more delicate thing than adding a scoop implies. You are not filling an empty tank. You are adding to a system that is already working, every second, to balance itself. Put the right amount of the right thing in at the right moment and you help that process along. Put in too much, or the wrong thing, or at the wrong time, and you’ve shoved a self-correcting system off the center it was trying to find — which is its own kind of harm. Are we helping it find balance, or knocking it further off?

Horse sweat is hypertonic: What that means and why it’s important

When a human sweats, the fluid that comes out is hypotonic — more dilute than blood. We lose proportionally more water than salt, which is why a person who only drinks plain water during heavy exercise can still do reasonably well, and why human sports drinks are mostly water with a light mineral content.

A horse is the opposite. Equine sweat is hypertonic — more concentrated in electrolytes than the horse’s own blood plasma. The horse loses proportionally more salt than water. This is not a small species footnote; it changes what replacement has to look like, and it changes how the horse’s own thirst behaves.

Because a horse sweats out a fluid saltier than its blood, heavy sweating can actually lower the sodium concentration of the blood rather than raise it. And it’s the rise in blood sodium that normally tells a body, in plain terms, that it’s thirsty. So a heavily sweating horse can be genuinely depleted and dehydrated and still not feel the urge to drink. The internal alarm that should be going off has been muffled by the very chemistry of equine sweat. Hold onto that, because it’s the hinge the whole “just make them drink” strategy turns on, and it’s exactly where that strategy gets dangerous.

Side Note: This mechanism of hypertonic sweat is also why the Gatorade trick doesn’t do what people think it does. Pouring a sports drink in the bucket replaces what a human loses — and a human loses a fraction of what a horse does, in a far weaker brine. A 20-ounce bottle of Gatorade carries about 270 milligrams of sodium. A single liter of horse sweat carries somewhere around 3,500 milligrams — roughly thirteen bottles’ worth — and a hard-working horse in heat can lose ten or more liters in an hour. One bottle replaces a negligible amount of what’s actually leaving, while delivering 34 grams of sugar to do it. In a horse’s bucket, Gatorade is essentially artificial flavoring and sugar, and if the whole goal is to get a horse to drink, there are better and healthier ways to encourage hydration.*Carefully climbs back down off soapbox*

What’s actually in the sweat

If sweat is the loss, we should be precise about what’s leaving. Across the research, equine sweat runs roughly in this range: sodium around 3.1 to 3.5 grams per liter, chloride around 6 grams per liter, potassium around 1.2 to 1.9 grams per liter, and then small amounts of calcium and magnesium — on the order of a tenth of a gram of each per liter, sometimes less.

Read that list again, because the proportions are the whole story. Chloride is the largest single loss. Sodium is second. Potassium is a distant third. Calcium and magnesium are trace. So if you’re going to replace what a horse loses, the thing you mostly need to put back is sodium chloride, which is to say: salt.

There’s one more thing I learned that is interesting so I’m going to share it. A meaningful fraction of horse sweat is protein — a surfactant called latherin, the reason equine sweat foams into lather between the hindquarters and under the tack. It’s what lets a horse wet its own coat for evaporative cooling despite all that hair. It’s a real loss, but it is not an electrolyte, and no electrolyte product replaces it. The horse rebuilds it from ordinary dietary protein.

Where the replacement actually comes from

Now that we understand what is being lost, let’s discuss what is added every single day in a balanced diet. Forage does the heavy lifting on potassium. Hay and pasture are naturally potassium-rich. A horse eating an appropriate amount of forage is, in almost all cases, meeting and exceeding its potassium requirement without anyone adding anything.(You can have your hay tested to verify and, if needed, adjust.)

A ration balancer or a complete feed covers the trace minerals, calcium, and magnesium — the calcium-to-phosphorus ratio, the copper, zinc, manganese, selenium, the things a forage-only diet tends to come up short on. But here is the sentence I need you to actually read, because it’s where most of these diets quietly fail: a ration balancer only delivers what it promises if it’s fed at the recommended rate.

A balancer is formulated to be fed at a specific daily amount — often a pound or two for an average horse — precisely because that’s the amount that carries a full day’s fortification. Feed less than that and the horse isn’t on a balanced diet anymore; it’s on a partially fortified one, and no amount of electrolyte supplement on competition day fixes a baseline that’s been quietly underfed for months. If you can’t or won’t feed a balancer at its recommended rate, that’s not a failure — it just means the baseline minerals need to come from somewhere else, deliberately, rather than being assumed.

What the diet does not reliably cover is sodium. This is the consistent hole. Forage is low in sodium. A ration balancer contains some salt, but fed at its small daily rate it contributes only a couple of grams of sodium — nowhere near a day’s requirement. This is not an accident or an oversight by the feed companies; it’s why nearly every balancer label tells you in plain print to provide free-choice salt alongside it. Sodium is the one macromineral the horse cannot store in meaningful reserve which is exactly why it’s the one we have to think about every single day.

What the published requirement actually is

Let’s put numbers to it, because “add salt” is vague and the vagueness is where people either under-do it or panic-buy. The National Research Council’s Nutrient Requirements of Horses is the reference the entire field works from, and for a roughly 1,100 pound horse at maintenance — standing around, light activity, temperate weather — the daily requirement lands near 10 grams of sodium, around 40 grams of chloride, and about 25 grams of potassium.

Forage hands you the potassium and a good share of the chloride. The sodium is on you. About two tablespoons of plain table salt — on the order of 34 grams — supplies roughly 13 grams of sodium and 20 grams of chloride, which covers the maintenance requirement and a bit of light, everyday sweat besides.

The requirement climbs with work, and the NRC scales it: moderate work pushes sodium toward the high teens of grams per day, heavy work into the high thirties, very heavy work past forty. But notice how it climbs — it climbs because the horse is sweating more, and what’s in that sweat is mostly sodium and chloride. The thing that goes up is, again, mostly the need for salt.

So when is it salt, and when is it electrolytes?

Here’s the decision, stripped down: An electrolyte supplement earns its place when the horse is losing trace minerals faster, and in greater total volume, than salt-plus-forage can comfortably replace — and specifically when the losses start to include enough potassium, calcium, and magnesium that plain salt no longer matches the pattern. That’s a real situation. It is just a narrower one than the shelf space suggests.

For the large majority of horses — light to moderate work, ordinary weather, a sound baseline diet — the honest answer is salt and water. Forage has the potassium handled. The balancer has the trace minerals handled. The sweat losses are modest enough that the extra sodium and chloride from a tablespoon or two of salt closes the gap. Reaching past that for a flavored electrolyte scoop isn’t dangerous in this case, but it’s usually replacing minerals the horse either already had or didn’t lose in meaningful quantity. That’s the “expensive pee” instinct I came in with.

But it is not true that electrolytes are always a waste. Sodium in particular isn’t stockpiled by the body the way some nutrients are, so a horse coming off a genuinely hard, sweaty effort does need active replacement — water alone won’t refill that tank, and in fact water alone after heavy sweat can dilute things further. The precise enemy isn’t “electrolytes.” It’s electrolytes given without matching the actual loss, and without making sure the horse drinks. Give a horse a big balanced electrolyte dose after a forty-minute schooling ride in spring and yes, most of it is going out the other end. Give that same dose to an endurance horse fifty miles into a hot ride and you’re replacing exactly what left. Same scoop. Completely different verdict. The question is never “electrolyte or not” in the abstract — it’s “does the loss in front of me actually call for this.”

The drinking problem — and why it’s the dangerous part

There is a caution to oversupplementation.

Remember how hypertonic sweat may trick a depleted horse into NOT feeling thirsty? A very common strategy has grown up around that gap — load the horse with salt or electrolytes specifically to drive thirst, to force the drink the horse isn’t volunteering. And on paper it works: raise the body’s sodium, trigger the thirst response, horse drinks, bucket empties.

The problem is what happens when the horse doesn’t drink. Now you’ve concentrated the horse further — pulled more salt into a body whose water tank you were trying to refill — and the horse is worse off than before you intervened. You’ve manufactured the dehydration you were trying to prevent. Giving a horse a concentrated electrolyte dose and hoping thirst follows is a bet, and the downside of losing that bet is a more dehydrated horse, not a neutral outcome.

So if the goal is genuinely that the horse drinks more water, the safer route isn’t to throw the chemistry off and gamble on thirst. It’s to make the water itself something the horse wants to drink — to work with the horse’s preferences rather than against its physiology.

When horses were offered flavored water alongside plain, voluntary intake did go up — but only for the flavor the horses actually liked. Horses were given four options: plain water, sweet-feed–flavored, apple-electrolyte–flavored, and peppermint. Of those four, the sweet-feed flavor beat plain water handily; horses drank meaningfully more of it, on the order of several gallons a day more for a full-sized horse. Plain peppermint and the apple-electrolyte flavor actually lost to plain water — the horses drank less of them than of plain.

The the most useful thing in the paper: flavor works, but the right flavor is individual, and the best way to find it by offering options. Of note, the three most appealing profiles for horses (fenugreek, banana and cherry) were not offered as options in that study.

What this looks like in the situations you’re actually in

The argument is only worth anything if it survives contact with real horses in real conditions. So here are the cases, sorted the way I think about them: the idle horse just trying to survive the weather, the horse traveling, and then the athlete across the disciplines. The thread running through all of them is the same — how much is actually leaving, and does that loss call for salt or for genuine electrolyte replacement.

That’s the spread. Two short-and-explosive disciplines where salt and water do most of the work, a middle band where balanced electrolytes start to matter on the hard hot days, and the long-haul athletes where serious replacement is genuinely required. The mistake I see — the one I came in biased against and still believe is the common one — is treating the barrel horse like the endurance horse. Buying the deep-end product for the shallow-end need.

Where I landed, after asking the science to argue with me

My bias mostly survived, but it came back more precise, and I think the detail is the useful part.

It’s true that salt is the thing most horses actually need which, in this economy, is a relief as it’s the cheapest thing on the shelf by a wide margin. It’s also true that forage and a properly fed ration balancer have already handled most of the rest.

It’s true that a great deal of electrolyte supplementation is replacing what wasn’t lost, and ends up as expensive urine — with the honest agreement in a need “that matches the loss.”

And it’s true that using electrolytes to force a non-drinking horse to drink is a risky proposition.

So the order of operations I’d hand anyone who asked is the same one I’ve built my own barn around. Feed the baseline diet at the rate it’s designed for. Add salt, every day, because the diet assumes you will. Make sure the horse has water it actually wants to drink — and find that out by offering choices, not by tipping its chemistry and gambling. And then, only when the work or the conditions are genuinely pulling more out of the horse than salt and forage can replace — real heat, real volume, real sustained sweat — reach for a real electrolyte, and match it to the loss in front of you.

And, because this is a topic where being wrong has real consequences, I am not your veterinarian, and I am not your nutritionist. What you’ve read here is me pushing back on my own bias and trying not to over- or under-supplement when it isn’t warranted — learning from the published requirements, what the science actually supports, and demonstrating how I think it through for my own horses. The ultimate authorities on your specific horse are the NRC for baseline requirements, your veterinarian for that animal’s health and any heat emergency, and a qualified equine nutritionist for dialing in the diet. Especially in extreme heat, in a horse that won’t sweat, or in a horse that won’t drink — that’s a phone call, not an article.

So start from what’s leaving the body and subtract what the diet already gives back… whatever’s left is the only thing you actually need to add. For most horses, on most days, that’s a couple of tablespoons of salt and a bucket they’re happy to drink from. The rest is for the horse that’s truly earning it — and on the hottest days, that horse might be standing in your own pasture.

References

• Walton, R.M., Cowell, R.L., & Valenciano, A.C. (eds.) (2020). Acid–Base and Electrolytes. In Equine Hematology, Cytology, and Clinical Chemistry, 2nd ed. Hoboken, NJ: Wiley-Blackwell. DOI 10.1002/9781119500186.ch7.
• Kerr, M.G., & Snow, D.H. (1983). Composition of sweat of the horse during prolonged epinephrine (adrenaline) infusion, heat exposure, and exercise. American Journal of Veterinary Research, 44(8): 1571–1577. PMID 6625308.
• McCutcheon, L.J., & Geor, R.J. (1998). Sweating: Fluid and ion losses and replacement. Veterinary Clinics of North America: Equine Practice, 14(1): 75–95. PMID 9561689.
• National Research Council (2007). Nutrient Requirements of Horses, 6th rev. ed. Washington, DC: The National Academies Press.
• Holbrook, T.C., Simmons, R.D., Payton, M.E., & MacAllister, C.G. (2005). Effect of repeated oral administration of hypertonic electrolyte solution on equine gastric mucosa. Equine Veterinary Journal, 37(6): 501–504. DOI 10.2746/042516405775314880. PMID 16295925.
• Van Diest, T.J., Kogan, C.J., & Kopper, J.J. (2021). The effect of water flavor on voluntary water intake in hospitalized horses. Journal of Equine Veterinary Science, 98: 103361. DOI 10.1016/j.jevs.2020.103361. PMID 33663710.
• Geor, R.J., McCutcheon, L.J., et al. (2000). Heat storage and thermoregulatory responses during exercise in hot/humid conditions in horses. (Equine exercise-physiology series; see also McCutcheon & Geor 1998, above.)
• Lindinger, M.I. (2022). Oral electrolyte and water supplementation in horses. Veterinary Sciences, 9(11): 626. DOI 10.3390/vetsci9110626.
• Waller, A.P., & Lindinger, M.I. (2021). Pre-loading large volume oral electrolytes: tracing fluid and ion fluxes in horses during rest, exercise and recovery. Journal of Physiology, 599(16): 3879–3896.
• Marlin, D.J., Scott, C.M., Roberts, C.A., Casas, I., Holah, G., & Schroter, R.C. (1998). Post exercise changes in compartmental body temperature accompanying intermittent cold water cooling in the hyperthermic horse. Equine Veterinary Journal, 30(1): 28–34. PMID 9458396.
• Johnson, E.B., MacKay, R.J., & Hernandez, J.A. (2010). An epidemiologic study of anhidrosis in horses in Florida. Journal of the American Veterinary Medical Association, 236(10): 1091–1097.
• Goodwin, D., Davidson, H.P.B., & Harris, P. (2005). Selection and acceptance of flavours in concentrate diets for stabled horses. Applied Animal Behaviour Science, 95(3–4): 223–232.

Follow the Money: What the Horse World Pays to Learn

The Science · The Library  ·  By Sara Kirkwood, Founder of Improve Equine  ·  June 2026

I went looking for proof that a more varied diet builds a more resilient horse. What I found instead was a map of who pays for equine science — and what those dollars are really asking.

I started with a question: if I deliberately built a diet around diversity — varied forages, a wider range of plants and fibers, the kind of mixed browsing a horse would do if left to its own devices — would I be able to create a resilient biome, one that would be less likely to be thrown by a sudden change in feed?

The logic is simple. For a gut to process a food, it has to have the bacteria on hand that can break that food down. So the theory runs like this: feed a varied diet and I’m feeding a diverse set of bacteria — and that range of microbes is the thing standing ready to process whatever range of inputs I put in front of my horse. Keep the bench deep, and a new feed stops being a shock to the system. It’s just one more thing the gut was already equipped to handle.

Because if the answer were yes, then the rule every one of us was raised on — introduce any new feed slowly, over ten to fourteen days, or risk colic — would start to look less like a fixed law of equine biology and more like a workaround for a fragile gut we built ourselves. That was the idea. I wanted to know if it had merit. So I went looking for the peer-reviewed research.

What I found was not a clean yes or a clean no. What I found was a hole.

The specific test — does feeding diversity build a biome resilient enough to absorb a change — has barely been run in horses. What exists sits just to the side of it, pointing the right direction without ever closing the loop. Horses given a choice of several forages instead of one forage more naturally and clearly prefer the variety. And as a broad principle, lower microbial diversity tends to mean lower gut stability. The mechanism is sound ecology — diversity in, diversity maintained, resilience as a result. But the horse study that would carry it from sound theory to demonstrated fact isn’t there. No one has run it.

And here is the part that turned a feeding question into a different kind of question entirely. The diet research that has been funded — the well-cited, repeated, peer-reviewed work — is almost all pointed the other direction. It is about managing the fallout of how we already feed. We know, from large prospective studies, that a change in batch of hay or type of grain raises the risk of colic, and that feeding large grain meals does measurable things to the gut. That’s good science. But look at what it’s for: it tells us how to safely operate the delicate system we designed. It does not ask whether we could build a sturdier one.

So I stopped asking about diet and started asking about money. Who funds equine science? What are they trying to prove? And who does it actually serve?

Who pays, and what they ask

The first name you hit, when you go looking for who pays for horse research, is the Grayson-Jockey Club Research Foundation — the best known, and the longest at it. It committed about $2.7 million in 2025, part of more than $44 million across 450-plus projects since 1940. It was built by Thoroughbred horsemen, its career-development award is named for a stallion, and while its findings genuinely benefit all breeds, the orientation shows in the portfolio: year after year, the largest categories of work are infectious disease and the injuries that end racing careers.

The Morris Animal Foundation is the other American pillar — more than $26 million across 670-plus equine studies since 1959 — and its record is built on vaccines, genetic screening tests, and disease.

In the UK the pattern isn’t subtle, it’s structural. The Horserace Betting Levy Board funds equine veterinary research from a statutory levy on horse-race betting — on the order of two to three million pounds a year, explicitly for the benefit of the Thoroughbred — and its stated priorities are disease control and the minimising of fatal and non-fatal injury in training and racing.

Breed associations follow suit. A recent American Quarter Horse Foundation round put $417,673 into projects on drug-resistant foal infections, osteoarthritis drugs, and wearable sensors — funding, in its own words, the “health, welfare and utility” of the horse.

Even the public money asks a production question. At the USDA’s research arm, horses are funded as livestock, inside programs for animal health and animal reproduction, where the named outcomes are vaccines, diagnostics, and reproductive efficiency.

And the flagship academic institute, the Gluck Equine Research Center at the University of Kentucky, is by its own description the only US center where nearly all faculty research equine health and diseases full-time — the birthplace of six of the ten major equine vaccines — sustained in large part by Thoroughbred breeding money whose custodians describe their mission as responding to existential threats to the breeding industry.

Then there’s the money that dwarfs all of the above: the pharmaceutical industry. The global animal-health market was already a $23 billion market when Zoetis — the world’s largest animal-health company, with 300-plus product lines — spun out of Pfizer. Boehringer Ingelheim runs on the order of $4.5 billion in annual animal-health research and development. And what is that money chasing, on the equine side? Their own catalog answers it: vaccines, de-wormers, joint therapy, and solutions for ulcers and pain. The shots. The joint injections. The ulcer meds. The painkillers.

The further I went, the more the pools looked alike. I kept expecting to find at least one big, disinterested source — money given to understand the horse for the horse’s own sake. Among the money that actually moves the field, I didn’t. Every major pool runs the same arithmetic: money in the front, a return out the back, the horse somewhere in the middle. Racing’s return is the race. Pharma’s is the cure. The feed company’s is the bag. The government’s is production. And the registry’s is the most literal of all — the foal itself, bred and registered and sold, then sold again into breeding and showing and racing, each its own market. There, the horse isn’t even the means to the product. The horse is the product. Almost none of it is spent without expecting the horse to earn it back.

What it buys the horse

The truth is, the research does help horses. Vaccines stop outbreaks that would kill them. Better diagnostics and surgical techniques and safer footing mean fewer catastrophic breakdowns. Genetic screens let us retire heritable disease out of the population. The merit is real and the benefit to horses is real, and nothing I’m turning over here takes that away.

But I keep coming back to the shape of it. Almost without exception, the money buys ways for the horse to tolerate us — to withstand being hauled and stabled in large groups, to keep running, to keep breeding, to keep performing on our schedule. Strip it all the way down and that is what we have funded, over and over: how to make a horse tolerate the things we do to it and the way we keep it. What we have rarely paid to ask is the other question entirely — whether the things we do, and the way we keep them, are part of what wears them down in the first place.

And I don’t think that’s a conspiracy, or even a failing. I think it’s history. For almost all of the time horses and humans have shared, a horse was livestock — a work tool, an asset, a way to get somewhere. The idea that we’d study a horse’s diet, or its gut, or its behavior for the horse’s own sake rather than for its output is genuinely recent. Species-appropriate care, gentler training, welfare treated as its own end — these are young conversations, and the money hasn’t caught up to them. Which is exactly why so much of what we know about them is still observational, or anecdotal, or built on a handful of small studies. The questions aren’t settled because no one has funded them long enough to settle.

What’s starting to change

Here is the budding shift, and the reason I think this is hopeful rather than bleak. That disinterested money — the kind given purely to understand the horse — does exist. It’s just small, and new. Welfare charities like World Horse Welfare and The Horse Trust fund research into equine quality of life and behavior for no reason other than the horse’s own good. The one dedicated equine-behavior fund I could find at a major foundation exists because a single donor created it, capped at $20,000 a project, after concluding such funding had been “minimal.” Twenty thousand dollars, beside billions. But it’s there. And around it, in just the last few years, a genuine body of peer-reviewed work has begun to build — led out of the UK, though international in reach — asking a different question than any of the big funders ever have. Not how to make the horse endure us. How the horse actually lives, what it feels, and what it is to us.

Some of it is aimed squarely at the bond. Researchers recently adapted the classic “strange situation” attachment test for horses and found that horses approach their own rider faster than a stranger and stay closer to them — behavioral evidence of a specific attachment, and an argument that we should be studying the socio-emotional side of the relationship at all (Kovács et al., 2026). Another team built the first validated scale to measure the human-horse bond, from more than 3,600 owners, and found it has six dimensions — companionship, personal wellbeing, dependence, status, growth, and, tellingly, sacrifice: the financial and personal investment owners pour in (Corrigan et al., 2025). Sacrifice didn’t show up as the opposite of love. It showed up as one of its measures. The money we spend on a horse and the bond we have with it are, it turns out, the same gesture.

Dog people have known this for a while, and the science followed them there. Because a dog’s whole value is the relationship, the funding flowed toward how dogs think, how they age well, how the bond works — an entire field of canine cognition and healthy-aging research, some of it federally funded, built around the life of the animal rather than its output. The horse world is arriving at that same door now. For more and more of us, the horse stopped being a tool a long time ago. The science is finally catching up to what we already felt.

We don’t have to burn it down to do better

None of this makes anyone a villain, and I want to be careful here. You can love a horse and still need it to earn its keep — horses are expensive, and there is no shame in that. The groom at the track who says “but we love our horses” is telling the truth. Loving a horse and seeing room to do better by it were never opposites; they live in the same person, easily. I know, because they live in me.

So here’s what I can’t stop thinking about. We have spent decades, and fortunes, funding ways to help horses tolerate how we keep them — chasing longevity, soundness, fewer ulcers, steadier minds, by managing the consequences. What if we spent even a fraction of that following the science that’s beginning to point upstream — toward better ways to feed them, house them, handle them, and start them? Not a teardown. Not throwing out everything we know. Small, evidence-based changes to how we already do things could move that bar — the same bar we’ve been straining to move from the wrong end — faster than another decade of buying the durability back one injection at a time.

A great deal of “the way we’ve always done things” is good. It’s hard-won and true and worth keeping. Modernizing it doesn’t mean discarding it; it means holding it up to the light of the research and the behavior of the horses themselves, and keeping what holds. That is the entire reason Improve Equine exists — not to tear down the way we’ve always done things, but to sort the wisdom from the habit, grounded in science and in what our horses are plainly trying to tell us. Keep the baby, just change the water. And maybe, if enough of us start asking the question with our attention and our dollars, the science will finally come meet us where the horses already are.

References

• Goodwin, D., Davidson, H.P.B., & Harris, P. (2002). Foraging enrichment for stabled horses: effects on behaviour and selection. Equine Veterinary Journal, 34(7), 686–691.
• Garber, A., Hastie, P., & Murray, J-A. (2020). Factors Influencing Equine Gut Microbiota: Current Knowledge. Journal of Equine Veterinary Science, 88, 102943.
• Tinker, M.K., White, N.A., Lessard, P., Thatcher, C.D., Pelzer, K.D., Davis, B., & Carmel, D.K. (1997). Prospective study of equine colic risk factors. Equine Veterinary Journal, 29(6), 454–458.
• Hudson, J.M., Cohen, N.D., Gibbs, P.G., & Thompson, J.A. (2001). Feeding practices associated with colic in horses. Journal of the American Veterinary Medical Association, 219(10), 1419–1425.
• Lopes, M.A.F., White, N.A., Crisman, M.V., & Ward, D.L. (2004). Effects of feeding large amounts of grain on colonic contents and feces in horses. American Journal of Veterinary Research, 65(5), 687–694.
• Kovács, A.B., Topál, J., & Gergely, A. (2026). Horses prefer their rider: A strange situation test adaptation shows behavioural evidence for partner-specific bonding. Journal of Equine Veterinary Science, 161, 105853.
• Corrigan, R.H., Pierard, M., Davies, E., Marlin, D., Evans, S., & Williams, J.M. (2025). Development of the Human-Equine Attachment Scale. Equine Veterinary Journal (advance online publication).

Related reading from The Library

• The Cost of “Free” Water
• Foundational Horse Nutrition, But Make It Make Sense