
Nutrition Science
Soaking hay: why no soaking time guarantees a result on sugar
Between 6 and 54 per cent: that is how widely sugar losses scattered when hay was soaked. What duration, water temperature and batch change.
HOHorseBmoderate Open access
MUMultiple speciesBmoderateT0not transferable
The induction dose in the fructan model is 7.5 grams per kilogram by stomach tube: converted, that is roughly twelve kilograms of dry matter grass at once.

Executive summary
The familiar fructan model of laminitis rests on single doses of 7.5 to 12.5 grams of oligofructose per kilogram of bodyweight given by nasogastric tube. Translated to pasture, the lowest of those doses would mean, for a 500 kilogram horse, roughly twelve kilograms of dry matter of the most fructan-rich grass ever measured, all at once. In parallel, raised insulin alone triggered laminitis in healthy horses under experimental conditions. Research has shifted its centre of gravity accordingly.
Key points
The idea that fructan in grass triggers laminitis does not come from pasture observation but from the laboratory. In 2006, twelve horses received oligofructose, an industrially produced form of fructan, as a single dose by nasogastric tube; six comparison animals were given a sham treatment. All twelve treated horses developed laminitis within 48 hours, detectable both clinically and in the tissue, and none of the comparison animals did. The closing sentence of that paper explicitly drew a link between cases seen in practice and the fructan content of pastures. Yet nothing had been measured on any pasture: the study had not watched a single horse graze.
Two years earlier the same question had been examined through a dose series. Three pairs of healthy horses received 7.5, 10 and 12.5 grams of oligofructose per kilogram of bodyweight, again as a single dose by tube. All developed laminitis after 48 hours, and as the dose rose the cells of the hoof lamellae increasingly lost their anchoring structures. That is the traceable origin of the figure of 7.5 grams per kilogram, which has circulated for years on manufacturers' pages without an author, without a year and without a reference, presented as a pasture risk.
No. It is the lowest dose tested in that series, given all at once through a tube. A lower value was never tested, so a minimum dose is unknown. The figure describes what reliably triggered laminitis in the experiment, not the point at which a pasture becomes dangerous.
How little robust the model is at its lower end was shown by a randomised trial from 2009. Three groups of eight horses there received endotoxin, oligofructose or both in sequence. With five grams of oligofructose per kilogram alone, only two of eight horses became laminitic; with the combination, five of eight; with endotoxin alone, none. Below the high doses, the model therefore no longer reliably triggers what it is meant to trigger.
These doses can be converted into feed as soon as a measured fructan content is placed alongside them. A British methods paper determined the fructan content of four pasture grasses between April and November using high-performance liquid chromatography. The values lay between 83 and 299 grams per kilogram of dry matter, with a mean of 154.
For a 500 kilogram horse, the lowest induction dose means 3.75 kilograms of pure fructan at once. At the highest content ever measured, that corresponds to roughly 12.5 kilograms of dry matter; at the mean content, to roughly 24 kilograms. For comparison: in a grazing trial, ponies with immediate access to the whole area took in on average 2.33 per cent of their bodyweight as dry matter per day, which is just under twelve kilograms for an animal of 500 kilograms. The arithmetic comes out the same across all bodyweights: at the mean content, the single dose would be about twice what an animal eats in a whole day.
Two objections belong with this. First, the widely used colorimetric method measures too low: on the same samples its mean was 82 rather than 154 grams per kilogram. Part of the publicly quoted fructan values therefore underestimates the true content by an unknown margin. Second, arithmetic says nothing about smaller amounts. That 3.75 kilograms at once trigger laminitis does not mean that a tenth of that remains without consequence. The calculation shows only how far the model stands from a pasture.
Alongside the fructan model, a second series of experiments tested something else: insulin on its own. In 2007, five healthy ponies with no history of laminitis received a continuous infusion that kept insulin high while blood glucose was deliberately held within the normal range; four randomly allocated comparison ponies did not receive it. All five treated animals developed laminitis in all four feet in less than 72 hours, on average after a good 55 hours. None of the comparison ponies fell ill. Digestive signs or indications of general illness were entirely absent.
In 2010 the same approach was repeated in sport horses. Four of four treated Standardbred trotters developed laminitis in less than 48 hours, none of the four comparison animals. All treated animals had a calculated insulin sensitivity in the normal range before the trial. The route is therefore not confined to ponies with disturbed metabolism.
In the experiment yes, but with an important qualification: the insulin values reached lay about 66 times above the animals' baseline and far above what a pony reaches at pasture. The experiments show that this route exists, not that it is taken in this form under pasture conditions.
A 2023 review sorts the findings. It separates three forms of laminitis: the endocrine form, the form accompanying severe general illness, and the form of the overloaded supporting limb. As a shared core process the authors propose stress of the basal cells of the hoof lamellae, which comes about differently in each of the three forms. For pasture laminitis they note that many affected ponies show the pattern of disturbed insulin regulation. That is a proposal by the authors, explicitly framed as a hypothesis, not a proven unifying theory.
If insulin is the lever, the question is which sugar pulls it. A randomised crossover trial in twelve ponies examined this directly, in spring and in autumn. Glucose, fructose and inulin, the commercially available form of fructan, were each fed in addition at one gram per kilogram of bodyweight per day. Inulin changed blood glucose and insulin only slightly, fructose distinctly less than glucose, and ponies with and without a history of laminitis separated most clearly after glucose. In autumn the insulin peaks were higher and more irregular than in spring.
An earlier paper reached a different result. There, insulin in laminitis-prone ponies rose to 5.5 times its starting value after inulin was fed, and those animals were not even overweight. The difference between the two papers probably lies in the amount: the earlier one used three grams per kilogram per day, three times as much as the later one. Taken together, both studies suggest that fructan could move insulin, but that this requires amounts far above the lower dose tested.
A further finding complicates the simple account in which fructan reaches the large intestine undigested and sends fermentation off the rails. A study of the stomach contents of 23 horses found the fructan concentration reduced by 84 per cent in grazing horses and by 54 per cent in hay-fed horses, above all in the non-glandular part of the stomach. In horses on a high-starch ration this breakdown was absent. A large part of the fructan therefore disappears before the large intestine, and how much depends on what the horse otherwise eats.
The practical question is usually: when is the sugar content highest? A 2022 review examined the available measurement series. Water-soluble carbohydrates mostly rise from morning to evening over the course of the day, but the measured range of that rise runs from minus 20 to plus 74 grams per kilogram of dry matter. There are therefore series in which the content was lower in the evening than in the morning. That is exactly where the mutually contradictory pieces of advice on this subject come from.
The author explicitly concludes that several factors must be considered together before an effect is attributed to any single measure. On top of that comes the measurement problem: fructan is only one part of the water-soluble carbohydrates, and the cheap colorimetric determination underestimates it, because the enzymes used do not fully break down the fructan of pasture grasses. The authors of that methods paper note that it cannot replace chromatography. Anyone reading a fructan value for a pasture is reading a figure with considerable uncertainty attached.
This cannot be answered robustly. The review states only that water-soluble carbohydrate contents tend to be higher at cool temperatures and in cold seasons. A study linking frosty nights to laminitis cases could not be found in this search. The advice therefore rests on plausibility, not on a measurement in horses.
Who falls ill can currently be described better than when. At a Finnish university clinic, 36 horses presented for laminitis were examined over sixteen months. In 89 per cent, indications of an underlying hormonal disorder were found. One third of those had pituitary dysfunction, two thirds a raised fasting insulin without a conspicuous coat. Among the animals with high insulin, 95 per cent showed outward features of being overweight. This is a snapshot from a single clinic, so it describes the make-up of the caseload, not a risk.
Prospective measurement was carried out in 446 British ponies aged seven and over, free of laminitis at entry and followed for three years. Low adiponectin and high insulin went together with the later occurrence of laminitis. After one year, laminitis had been diagnosed by a vet in 4.0 per cent of the animals, after two years in 6.7 and after three years in 9.9 per cent. At the start, 72.2 per cent of the ponies were overweight or obese. The markers were therefore measurable before the first episode, although each value was determined only once.
Pituitary dysfunction is more common than was long assumed. A systematic review from 2018 gives as its most robust estimates 21.2 per cent in horses and ponies aged fifteen and over and 2.9 per cent in the overall population. Laminitis is among the most frequently reported signs there, alongside coat changes and muscle loss over the back. Increasing age remained the only established risk factor.
Conversely, the epidemiology of naturally occurring laminitis is thin. A systematic review screened 17 publications and rated six as the most reliable. Only the association between increasing age and chronic laminitis rested on good evidence; for sex, breed and bodyweight the results remained contradictory.
One quantity is missing from almost every discussion of pasture and sugar: the amount taken in. A trial in twelve ponies over 28 days put a number on it. The individual plots were sized to yield 1.5 per cent of bodyweight as dry matter per day. The ponies with immediate access to the whole area took in a calculated 2.33 per cent, distinctly more than intended, and gained 4.82 per cent of their weight in four weeks.
In the two groups given the area in portions behind an electric fence, the values were 1.59 and 1.82 per cent, and weight gain 1.16 and 1.54 per cent. Body condition score and the assessment of the neck crest rose only in the group with immediate full access. Laminitis was not an endpoint in this trial, and intake was calculated from the growth of the plot rather than measured in the individual animal.
The chain running from rapidly fermentable carbohydrates through acidification of the gut to laminitis has also been studied in cattle, where rumen acidosis has been centre stage for decades. A systematic review from 2023 included sixteen papers out of 339 screened. Fifteen of them confirmed the occurrence of rumen acidosis, yet in two papers that used oligofructose for the purpose not a single clinical sign of laminitis appeared. The chain is therefore not automatic even where it has been studied longest.
The inflammatory side also carries less weight than was long assumed. In one trial, twelve of 22 ponies received a ration with around 42 per cent non-structural carbohydrates for seven days, against around seven per cent in the comparison ration. In the hoof lamellae none of the pro-inflammatory messengers measured rose, and white blood cells did not migrate in. Only one particular enzyme was read out in greater amounts. The authors conclude from this that the marked lamellar inflammation of the sepsis models may play no central role in pasture laminitis.
Finally, nobody knows exactly how common laminitis is. A systematic review appraised 69 publications, kept ten as the most reliable and found frequencies between 1.5 and 34 per cent. The authors judge the quality of the papers to be predominantly poor and consider comparison of these figures with one another explicitly inadmissible, because the studies look at different causes of laminitis. Anyone reading a percentage for the frequency of pasture laminitis is reading a figure without a secure basis.
Original analysis
| Bodyweight | Single fructan dose in the model | Dry matter needed at the highest content (299 g/kg) | At the mean content (154 g/kg) | At the lowest content (83 g/kg) | Measured daily intake with full access |
|---|---|---|---|---|---|
| 200 kg | 1.5 kg | 5.0 kg | 9.7 kg | 18.1 kg | 4.7 kg |
| 300 kg | 2.25 kg | 7.5 kg | 14.6 kg | 27.1 kg | 7.0 kg |
| 400 kg | 3.0 kg | 10.0 kg | 19.5 kg | 36.1 kg | 9.3 kg |
| 500 kg | 3.75 kg | 12.5 kg | 24.4 kg | 45.2 kg | 11.7 kg |
| 600 kg | 4.5 kg | 15.1 kg | 29.2 kg | 54.2 kg | 14.0 kg |
| 700 kg | 5.25 kg | 17.6 kg | 34.1 kg | 63.3 kg | 16.3 kg |
The data point to no generally valid time of day. A 2022 review summarises that water-soluble carbohydrates mostly rise from morning to evening over the course of the day, but that the measured range of that rise runs from minus 20 to plus 74 grams per kilogram of dry matter. In individual measurement series the content was therefore lower in the evening than in the morning. Higher contents go together with cool temperatures and cold seasons, frequent grazing lowers them, and nitrogen fertilisation and drought stress act inconsistently. The author therefore explicitly advises against deriving a rule from any single factor.
That figure is not known, and the number in circulation answers a different question. It comes from a dose series in which healthy horses received 7.5, 10 or 12.5 grams of oligofructose per kilogram of bodyweight once by nasogastric tube, in order to trigger laminitis deliberately. A lower step was never tested, so a minimum dose is unknown. In another experiment, five grams per kilogram were enough in only two of eight horses. These are experimental doses used to trigger the disease, not a limit value for a pasture.
It is established that very high single doses by tube trigger laminitis; it is not established that grazed grass does the same. For a 500 kilogram horse, the lowest experimental dose would correspond to roughly 12.5 kilograms of dry matter of the most fructan-rich grass ever measured, all at once, which is more than one measured daily intake. On top of that, the fructan concentration in the stomach of grazing horses was reduced by 84 per cent in one study. The focus of research has shifted to insulin accordingly. A cleanly measured link between pasture content and disease is still missing.
This cannot be answered with the available data. The review of sugar contents states that these tend to be higher at cool temperatures and in cold seasons. A study linking frosty nights to actual laminitis cases could not be found in this search. The widespread advice therefore rests on a plausible chain, not on a measurement in horses. For a horse whose metabolic situation is known, the attending veterinary practice is the right point of contact, because on current data the risk sits more in the animal than in the grass.
There is no established figure for this, and this article deliberately names none. The specialist literature contains no trial that compared different spring turnout plans and measured laminitis as an endpoint. Something else was measured: in a trial of twelve ponies over four weeks, the animals with immediate access to the whole area took in 2.33 per cent of their bodyweight as dry matter per day instead of the intended 1.5 per cent, and 1.59 to 1.82 per cent where the area was allocated in portions. How an individual horse is managed is decided by the attending veterinary practice.
Yes, being overweight is neither a precondition nor a proof. In one feeding trial, ponies that were explicitly not overweight but prone to laminitis showed distinctly reduced insulin sensitivity, which only became apparent when the metabolism was challenged. Conversely, in a Finnish clinical series 95 per cent of the horses with raised insulin showed outward features of being overweight, and in a British cohort 72.2 per cent of the ponies were overweight or obese at entry. Being overweight is therefore common, but not necessary. Whether a particular animal is affected can only be established by a veterinary examination.
Terms defined in this text

Nutrition Science
Between 6 and 54 per cent: that is how widely sugar losses scattered when hay was soaked. What duration, water temperature and batch change.
HOHorseBmoderate Open access

Equine Health
In autumn ACTH rises in healthy horses too: where the 29 and 47 pg/ml limits come from, and what breed, age and laboratory shift.
HOHorseArobust Open access

Equine Health
Bilateral forelimb lameness separates laminitis from other lameness most sharply, the digital pulse follows: what the studies actually show.
HOHorseBmoderate Open access
Once a month: new syntheses, checked studies, and what has changed in the state of the evidence.