Crib-biting and weaving: what the evidence says about stereotypies in horses

What crib-biting and weaving are, and how common they are

Stereotypies are repeated, uniform sequences of movement with no recognisable goal. In crib-biting, the horse places its upper incisors on a fixed object, tenses the muscles of the lower neck and emits a guttural sound as it does so. In weaving, it shifts its weight rhythmically from one forelimb to the other while standing and swings its head and neck from side to side. Box-walking describes pacing the same route around the box over and over. Research usually analyses the three patterns together, because surveys distribute them in a similar way.

How common this is can be answered for Switzerland with a single figure. A representative written survey recorded 2,536 horses from 622 yards. For 418 of them, or 16.5 per cent, the yards described an abnormal behaviour. Crib-biting, weaving and box-walking together affected 89 horses, or 3.5 per cent. Other abnormal behaviour such as pawing, tongue-play, head-shaking or kicking the box wall accounted for 11.1 per cent. These figures date from 2002 and have not been updated since.

Is crib-biting more common in certain breeds?

In a US survey the overall frequency of crib-biting was 4.4 per cent, and Thoroughbreds were the breed most affected. The research group therefore considers a genetic predisposition worth examining, and not the housing system alone. A reliable breed-specific frequency cannot, however, be drawn from that work.

When stereotypies begin: the trail leads to weaning

The most precise description of onset comes from a prospective study that followed 225 young Thoroughbreds and Thoroughbred crosses over four years. A good third of these horses, 34.7 per cent, developed an abnormal behaviour during that time. The individual patterns did not appear at the same time, but in a recognisable order.

One feeding characteristic stood out: where concentrate feed was given after weaning, the rate at which new crib-biting appeared was more than four times as high. Two further associations concerned abnormal behaviour as a whole and not crib-biting in particular: weaning with stabling rather than at pasture went with a rate more than twice as high, and housing in a building rather than at grass with a two-and-a-half-fold rate. Foals of high-ranking mares developed abnormal behaviour more often than foals of middle- or low-ranking mares.

An early review from 1999 puts this in context: the majority of stereotypies in the horse begin in the month after weaning, when the foal's feeding and social environment changes fundamentally. The same work considers the widespread explanation of boredom too general to be of any use in the yard.

These figures describe a single breeding population, and they describe onset, not disappearance. Every association found runs from rearing practice to the beginning of the behaviour. About a pattern that is already established this study says nothing.

Copying: the assumption the data contradict most clearly

In a US-wide owner survey, 54.4 per cent of respondents held environmental conditions to be the dominant cause of crib-biting, and 48.8 per cent assumed that horses learn the behaviour from one another by observation. The same survey also asked how the behaviour had actually begun in the affected horses. In 1.0 per cent this happened after the horse had encountered another horse that crib-bit.

The second finding is just as clear. 86 per cent of the horses recorded were at pasture with other horses, and the extent of contact with other horses was statistically not linked to risk. The research group concludes that there is little to support copying and that isolating affected horses can cause unnecessary stress.

Why the assumption holds nonetheless is shown by a qualitative interview study. Those who had never looked after a crib-biting horse named copying as a strong theme and did not want such a horse. Those who had one did not believe in it on the strength of their own observation, and would have bought or looked after another. A review from 1999 reached the same conclusion even then: the anecdotal reports of copying have neither direct nor indirect support.

The consequences of this assumption are measurable. A telephone survey of 300 English yards with 8,427 horses between them showed that 30 to 48 per cent of yards kept affected horses apart from the rest. In riding schools, 32 per cent would not take such horses in at all, in competition yards 17 per cent, in racing yards 4 per cent.

What anti-crib collars, muzzles and surgery actually do

The methodologically strongest trial on anti-crib devices is a randomised crossover trial in eight established crib-biters. They wore two different anti-crib collars and a muzzle one after the other, each device for seven days, with seven days without in between. All three clearly lowered the frequency of crib-biting compared with the comparison periods, and none was better than the others. The behaviour did not stop as a result: it was prevented while the device was worn.

Does a horse crib-bite more once the device comes off?

On this point two randomised trials contradict each other. In the crossover trial of 2016, the day after removal differed neither from the time before fitting nor from the comparison periods. An older trial from 1998, in which eight horses wore the collar for only 24 hours, found clearly more crib-biting on the following day than before.

The 1998 research group read this rise as a sign that the internal motivation increases during prevention, and expressly rejected two obvious competing explanations. Both trials involve eight horses, both measure only a single day after removal, and the wearing time differs by a factor of seven. The question is therefore open, not settled.

For the surgical route there is a case series of 119 horses operated on between 1994 and 2012. Of the 90 animals followed up, 76, or 84.4 per cent, had not crib-bitten for more than a year. 14 began again, on average after half a year. Where crib-biting had been present for more than three years before the procedure, the prospect of success was clearly lower. In 22.2 per cent complications occurred: haematomas, accumulations of serous fluid, infections or persistent wound discharge.

This series has no comparison group, follow-up is missing for 29 of the 119 horses, and what was measured was solely the absence of the movement. Whether the horse is better off afterwards was not recorded: neither welfare, nor remaining motivation, nor colic risk.

Stress, cognition and sensitivity: what stereotypies say about the horse

One trial compared 22 crib-biting with 21 non-crib-biting horses under a pharmacologically triggered stress response. Heart rate and heart rate variability did not differ. Salivary cortisol was on average higher in the crib-biters. The split within that group is instructive: crib-biters that did not crib-bite during the three hours had clearly higher values than the comparison horses; crib-biters that did crib-bite did not.

The research group accordingly reads crib-biting as a coping strategy that helps lower the cortisol level in demanding situations, and concludes that preventing the behaviour would then be inappropriate. The spread urges caution: the means lie close together, the value ranges overlap largely, and the groups were defined by a pre-existing characteristic rather than allocated at random.

Two further pieces of work clear away side assumptions. In four learning tasks, two of them reversal tasks, crib-biting horses needed as many trials to reach the learning criterion as the comparison horses and showed the same circulatory responses. Stereotypic behaviour therefore does not go with poorer learning performance. The research group does note, however, that the horses were deliberately not prevented from crib-biting during the tasks.

In a standardised test battery with 19 crib-biting and 18 non-crib-biting horses, the groups did not differ in reactivity towards people, in social behaviour, in locomotor activity or in curiosity. Only one characteristic stood out: crib-biters reacted more sensitively to touch. A consistently different personality profile does not follow from that, and the direction of the relationship remains open.

Stomach and gut: the physical trail

A trial in 35 young horses compared 19 animals that had recently begun to crib-bite with 16 unaffected animals over 14 weeks. At the first gastroscopy the stomach lining of the crib-biting foals was clearly more damaged and more inflamed. The acidity of the faeces was also lower, at a pH of 6.05 against 6.58.

After this examination the animals were allocated at random to an acid-buffering or a control ration. The buffering ration clearly improved the stomach findings. Crib-biting, however, declined in most foals regardless of the feeding, with only a trend towards a stronger decline on the buffering ration. What is established, then, is a link with the stomach and an effect of the ration on the stomach, but not a disappearance of the behaviour through treating the stomach.

Does a diseased stomach cause crib-biting?

That is not settled. The study shows that young crib-biters had more damaged stomachs; it does not show what came first. The inflammation may precede the behaviour or follow it, not least because crib-biting is suspected to influence salivation and gut passage. Both are compatible with the data available.

Colic: the best-documented health risk

The strongest link between crib-biting and a disease concerns a rare but severe form of colic: entrapment of small intestine in the epiploic foramen, a narrow natural opening in the upper abdomen. In an international case-control study with 109 affected and 310 comparison horses, crib-biting was by far the strongest factor. The margin of uncertainty around the estimate, however, runs from about 15-fold to nearly 300-fold: that a link exists is therefore certain, how large it is, is not.

An older analysis at two university clinics shows the same direction and, at the same time, how unstable the magnitude remains. In Illinois, 13 of 19 affected horses had a history of crib-biting against 2 of 34 horses with other strangulating small intestinal lesions. In Liverpool it was 24 of 49 against 72 of 687 horses with colic of another cause. The estimated magnitudes differ between the two clinics by more than fourfold.

How often colic occurs in crib-biting horses at all was examined by a British survey of 367 crib-biting horses from the general horse population. 130 of them, or 35.4 per cent, had at least one colic in the course of their lives, 672 episodes in total, 13 of them with surgery. In the preceding twelve months, 67 of 331 continuously supervised horses had a colic. Linked to the risk were, among others, longer time stabled in autumn and a behaviour triggered by feed intake.

This survey lacks a comparison group of non-crib-biting horses: within the study the 35.4 per cent can be compared with nothing. A case-control study of recurrent colic supplies that comparison. There, crib-biting went with an around tenfold higher chance of a further colic, weaving with a not quite fourfold one. For weaving this is one of the few published figures on a health consequence, and it comes from a single model resting on 59 cases.

Housing and feeding: what the observational data suggest

The analysis relevant to Switzerland examined 60 possible risk factors in 2,341 horses from 622 yards. Eleven showed up when considered singly, three remained in the joint analysis: adult warmbloods and Thoroughbreds described as reactive by the people looking after them, feeding in four meals a day, and no daily turnout.

What the research group derives from this it expressly labels a hypothesis: prevention that addresses causes would have to rely on housing conditions permitting bodily contact between horses, on daily free movement, and on plenty of forage with little or no concentrate feed. These assumptions were to be tested experimentally or by further observational studies. To this day they have not been.

A review from 2021 analysed work in three languages and found a very wide spread in the reported frequencies from one study to the next. Between surveys and direct behavioural observation there was no meaningful difference in this respect. As the most important influence on onset and frequency it names unsuitable feeding management. It gives no averaged frequency.

On feeding technique there is a semi-systematic overview of 26 pieces of work published between 2009 and 2025. Haynets and slow feeders lengthened eating time by around 40 per cent depending on mesh size and forage type, and lowered feed waste, in the best case, from 57 to 6 per cent. A decline in crib-biting and weaving is reported there as an observed side effect, qualitatively and without any effect size. Devices that restrict too strongly, by contrast, triggered frustration, and nets hung high increased the tension in the neck muscles.

Five widespread claims about crib-biting, each set against the strongest study design found

Original analysis

Limitations and uncertainty

• There is no systematic review with pooled effect estimates on crib-biting and weaving. The overview of 2021 rests on descriptive statistics and on data analysis, without a formal assessment of the risk of bias. The editorial requirement to base at least half the sources at the level of systematic review, meta-analysis or randomised trial cannot be met in this field: only four of the works used here reach that level.
• The behavioural trials involve six to a good twenty horses per group. The two randomised trials on anti-crib devices reach opposite results at different wearing times, and both measure the period after removal for a single day only.
• The only population-based frequency figure for Switzerland comes from a written survey of 2002 whose response rate barely exceeded a third. The two Swiss publications rest on the same survey and are not evidence independent of each other.
• The colic figures come from clinic and survey populations in which the behaviour was reported after the fact by the people looking after the horses. The estimated magnitudes vary considerably between study sites, and the survey with the largest number of crib-biting horses has no comparison group at all.
• The surgical case series has no comparison group, follow-up is missing for around a quarter of the horses operated on, and only the absence of the movement counted as success. Neither welfare nor remaining motivation nor colic risk was recorded after the procedure.
• Almost all the associations with housing and feeding come from observational studies. None of the factors named has so far been tested as a preventive measure in a controlled trial, and the rearing data concern a single Thoroughbred breeding population.

Open questions

• Does the motivation to crib-bite increase during mechanical prevention, and does the result depend on how long the device is worn?
• Does the raised sensitivity to touch in crib-biting horses precede the behaviour, or is it a consequence of long-standing strain?
• Do more forage, daily free movement and bodily contact between horses lower the frequency of stereotypies when they are deliberately changed in a controlled trial?
• Does the stomach damage in young crib-biters precede the behaviour, or does it arise only with it?

Frequently asked questions

Can my horse pick up crib-biting from another horse?

There is little to support that. In a US-wide owner survey, 48.8 per cent of respondents held crib-biting to be something horses copy, yet in only 1.0 per cent of the affected horses had the behaviour begun after the animal had encountered a crib-biter. 86 per cent of the horses were at pasture with others, and the extent of contact with other horses was not linked to risk. A review from 1999 found neither direct nor indirect support for the anecdotal reports. Imitation in an individual case is not ruled out by this, but as a major route of onset it is not supported.

Should I keep a crib-biter apart from the other horses?

The data argue against it. The research group behind the US survey expressly states that isolating affected horses can cause unnecessary stress, because there is little to be found in support of copying. In practice the opposite happens: in a survey of 300 English yards, 30 to 48 per cent kept affected horses apart, and depending on the type of yard, 4 to 32 per cent would not take such horses in at all. Whether and how a particular horse is housed nonetheless belongs to the judgement made on site, not to a general rule.

Is an anti-crib collar any use?

It suppresses the behaviour for as long as it is worn. In a randomised crossover trial in eight established crib-biters, two different collars and a muzzle clearly lowered the frequency of crib-biting, without the devices differing from one another. Whether the frequency rises again after removal is open: the trial of 2016 found no such rise after seven days of wearing, a trial of 1998 did find one after 24 hours. The research group of 2016 itself states that the effort would be better spent on the housing conditions than on the device.

Can a horse be broken of crib-biting?

No demonstrated procedure that dissolves an established crib-biting habit was to be found in this research. All the data on onset come from rearing and run from housing practice to the beginning of the behaviour, not to its disappearance. The devices tested prevent the behaviour while they are worn. The only intervention with lasting figures is surgical, and its measure of success is limited to the absence of the movement. What went with fewer stereotypies in the observational data concerns forage, free movement and social contact: none of it has been tested as a preventive measure.

Is crib-biting dangerous for my horse's health?

A link with colic is established several times over; its magnitude is not. It is clearest for entrapment of small intestine in the epiploic foramen, a rare but severe form of colic. In a survey of 367 crib-biting horses, 35.4 per cent had at least one colic in the course of their lives, but without a comparison group. For recurrent colic, crib-biting went with an around tenfold higher chance. Whether stopping the behaviour alters that risk has never been examined, and no study has measured the colic risk after surgery.

Is weaving just as problematic as crib-biting?

There are considerably fewer data on weaving. In the prospective rearing study, 4.6 per cent of the young horses started to weave, at a median age of 60 weeks, so appreciably later than for crib-biting at 20 weeks. On health consequences there is essentially a single figure: in a case-control study of recurrent colic, weaving was associated with a not quite fourfold higher chance, crib-biting with an around tenfold one. That figure comes from a single model resting on 59 cases and allows no sound comparison between the two behaviour patterns.

Are crib-biting horses more difficult or less able to learn?

No, no sign of that was found. In four learning tasks, two of them reversal tasks, crib-biting horses needed as many trials as the comparison horses to reach the learning criterion and showed the same circulatory responses. In a standardised test battery with 19 crib-biting and 18 non-crib-biting horses, reactivity towards people, social behaviour, locomotor activity and curiosity did not differ. The only thing that stood out was a greater sensitivity to touch. One restriction applies: in the learning tasks the horses were deliberately not prevented from crib-biting, so how they perform under prevention remains unknown.

Sources

1. Albright JD, Mohammed HO, Heleski CR, Wickens CL. Crib-biting in US horses: breed predispositions and owner perceptions of aetiology. Equine Veterinary Journal, 2009 (Cross-sectional study | Horse)DOI 10.2746/042516409x372584
   At an overall crib-biting frequency of 4.4 per cent, 48.8 per cent of owners held the behaviour to be something horses copy, while only 1.0 per cent of horses began after contact with a crib-biter and the extent of contact with other horses was not linked to risk; the research group concludes that isolation causes unnecessary stress and that a genetic predisposition remains worth examining.
2. McBride SD, Long L. Management of horses showing stereotypic behaviour, owner perception and the implications for welfare. The Veterinary Record, 2001 (Cross-sectional study | Horse)DOI 10.1136/vr.148.26.799
   In 300 English yards with 8,427 horses, 67 to 79 per cent prevented the behaviour mechanically while only 35 to 43 per cent addressed the conditions that trigger it, and 30 to 48 per cent kept affected horses apart; the concern was performance, clinical consequences and market value.
3. Litva A, Robinson CS, Archer DC. Exploring lay perceptions of the causes of crib-biting/windsucking behaviour in horses. Equine Veterinary Journal, 2010 (Other | Horse)DOI 10.1111/j.2042-3306.2009.00025.x
   Qualitative interview study: those who had never looked after a crib-biting horse named copying as a strong theme and did not want such a horse, while those who had one did not believe in it on the strength of their own observation; the dominant explanations were boredom, stress and habit.
4. Nicol C. Understanding equine stereotypies. Equine Veterinary Journal. Supplement, 1999 (Other | Horse)DOI 10.1111/j.2042-3306.1999.tb05151.x
   Narrative review: the boredom explanation is too general, the majority of stereotypies begin in the month after weaning, copying between horses has neither direct nor indirect support, and surgical or pharmacological prevention should not be attempted before the underlying causes have been removed.
5. Waters AJ, Nicol CJ, French NP. Factors influencing the development of stereotypic and redirected behaviours in young horses: findings of a four year prospective epidemiological study. Equine Veterinary Journal, 2002 (Cohort study | Horse)DOI 10.2746/042516402776180241
   Of 225 young horses followed prospectively, 34.7 per cent developed an abnormal behaviour, crib-biting 10.5 per cent at a median of 20 weeks and weaving 4.6 per cent at 60 weeks; concentrate feed after weaning went with a more than fourfold raised rate of newly appearing crib-biting, and simple changes to feeding, housing and weaning could lower the frequency substantially.
6. Bachmann I, Audige L, Stauffacher M. Risk factors associated with behavioural disorders of crib-biting, weaving and box-walking in Swiss horses. Equine Veterinary Journal, 2003 (Cross-sectional study | Horse)DOI 10.2746/042516403776114216
   Of 60 risk factors examined in 2,341 Swiss horses, three remained: adult warmbloods and Thoroughbreds described as reactive, four meals a day, and no daily turnout; the causal prevention derived from this, through bodily contact, free movement and plenty of forage, is expressly labelled by the authors as a hypothesis still to be tested.
7. Bachmann I, Stauffacher M. [Prevalence of behavioral disorders in the Swiss horse population]. Schweizer Archiv fur Tierheilkunde, 2002 (Cross-sectional study | Horse)DOI 10.1024/0036-7281.144.7.356
   In a representative Swiss survey of 2,536 horses, 16.5 per cent showed an abnormal behaviour and 3.5 per cent crib-biting, weaving or box-walking; the title is the translation indexed by PubMed, the work itself appeared in German and rests on the same survey as the risk factor analysis of 2003.
8. Seabra JC, Dittrich JR, Vale MMD. Factors Associated With the Development and Prevalence of Abnormal Behaviors in Horses: Systematic Review With Meta-Analysis. Journal of Equine Veterinary Science, 2021 (Meta-analysis | Horse)DOI 10.1016/j.jevs.2021.103750
   The overview finds a very wide spread in the reported frequencies, no meaningful difference between survey and direct observation, and names unsuitable feeding management as the most important influence, while resting on descriptive statistics and data analysis rather than on pooled effect estimates.
9. Albright JD, Witte TH, Rohrbach BW, Reed A. Efficacy and effects of various anti-crib devices on behaviour and physiology of crib-biting horses. Equine Veterinary Journal, 2016 (Randomised trial | Horse)DOI 10.1111/evj.12534
   In a randomised crossover trial in eight horses, two anti-crib collars and a muzzle lowered crib-biting with no difference between them and with no detectable rise on the day after removal, while surgically placed gum rings in six horses worked only temporarily and were probably painful; the research group calls for the housing conditions to be addressed rather than the behaviour prevented mechanically.
10. McGreevy PD, Nicol CJ. The effect of short-term prevention on the subsequent rate of crib-biting in thoroughbred horses. Equine Veterinary Journal. Supplement, 1998 (Randomised trial | Horse)DOI 10.1111/j.2042-3306.1998.tb05142.x
    An anti-crib collar prevented crib-biting in six of eight horses over 24 hours, and on the following day these horses crib-bit clearly more often than the comparison horses and than before, which the authors read as an increase in internal motivation during prevention.
11. Nicol CJ, Davidson HP, Harris PA, Waters AJ. Study of crib-biting and gastric inflammation and ulceration in young horses. The Veterinary Record, 2002 (Randomised trial | Horse)DOI 10.1136/vr.151.22.658
    19 young horses that had begun crib-biting shortly before had clearly more damaged and more inflamed stomachs at the outset than 16 unaffected animals, together with a lower faecal pH; a randomly allocated acid-buffering ration improved the stomach findings, but crib-biting declined in both groups, only somewhat more strongly on that ration.
12. Briefer Freymond S, Bardou D, Briefer EF, Bruckmaier R. The physiological consequences of crib-biting in horses in response to an ACTH challenge test. Physiology & Behavior, 2015 (Controlled trial | Horse)DOI 10.1016/j.physbeh.2015.07.015
    In 22 crib-biting and 21 non-crib-biting horses, heart rate and variability did not differ, salivary cortisol was on average higher in the crib-biters, and only those crib-biters that did not crib-bite during the test showed clearly higher values than the comparison horses: the authors read crib-biting as a coping strategy.
13. Briefer Freymond S, Ruet A, Grivaz M, Fuentes C. Stereotypic horses (Equus caballus) are not cognitively impaired. Animal Cognition, 2019 (Controlled trial | Horse)DOI 10.1007/s10071-018-1217-8
    In four learning tasks including two reversal tasks, crib-biting and non-crib-biting horses needed as many trials to reach the learning criterion and showed the same circulatory responses, the horses having deliberately not been prevented from crib-biting.
14. Briefer Freymond S, Bardou D, Beuret S, Bachmann I. Elevated Sensitivity to Tactile Stimuli in Stereotypic Horses. Frontiers in Veterinary Science, 2019 (Case-control study | Horse)DOI 10.3389/fvets.2019.00162
    19 crib-biting and 18 non-crib-biting horses did not differ in reactivity towards people, in social behaviour, in locomotor activity or in curiosity, only sensitivity to touch being raised in the crib-biters; the authors do not derive a consistently different personality profile from this.
15. Archer DC, Pinchbeck GK, French NP, Proudman CJ. Risk factors for epiploic foramen entrapment colic: an international study. Equine Veterinary Journal, 2008 (Case-control study | Horse)DOI 10.2746/042516408X266079
    In 109 affected and 310 comparison horses, crib-biting was by far the strongest factor for entrapment in the epiploic foramen, albeit with a very wide margin of uncertainty; the authors stress that the causal pathway remains unexplained and that it is unresolved whether a change of behaviour lowers the risk.
16. Archer DC, Freeman DE, Doyle AJ, Proudman CJ. Association between cribbing and entrapment of the small intestine in the epiploic foramen in horses: 68 cases (1991-2002). Journal of the American Veterinary Medical Association, 2004 (Case-control study | Horse)DOI 10.2460/javma.2004.224.562
    At two university clinics, 13 of 19 and 24 of 49 affected horses respectively had a history of crib-biting against 2 of 34 and 72 of 687 comparison horses, the estimated magnitudes differing between the two sites by more than fourfold; the authors state the link expressly as a possibility.
17. Escalona EE, Okell CN, Archer DC. Prevalence of and risk factors for colic in horses that display crib-biting behaviour. BMC Veterinary Research, 2014 (Cross-sectional study | Horse)DOI 10.1186/1746-6148-10-S1-S3
    Of 367 crib-biting horses, 130, or 35.4 per cent, had at least one colic in the course of their lives, and linked to this were, among others, longer time stabled in autumn as well as a behaviour triggered by feed intake; a comparison group of non-crib-biting horses is missing.
18. Scantlebury CE, Archer DC, Proudman CJ, Pinchbeck GL. Management and horse-level risk factors for recurrent colic in the UK general equine practice population. Equine Veterinary Journal, 2015 (Case-control study | Horse)DOI 10.1111/evj.12276
    In an analysis nested within a cohort, with 59 recurrences and 177 comparison horses, crib-biting went with an around tenfold and weaving with a not quite fourfold raised chance of a further colic, while more time at grass was only very weakly protective.
19. Amaje J, Upton S, Garba UM, Jolayemi KO. The use of hay nets and slow feeders as feeding methods in horse management: A semi-systematic review. Journal of Equine Veterinary Science, 2026 (Other | Horse)DOI 10.1016/j.jevs.2025.105762
    From 26 pieces of work published between 2009 and 2025: haynets and slow feeders lengthen eating time by around 40 per cent and clearly lower feed waste, a decline in crib-biting and weaving is reported qualitatively, while devices that restrict too strongly or hang too high cause frustration and more tension in the neck muscles; the work is declared semi-systematic, without any assessment of the risk of bias.
20. Baia P, Burba DJ, Riggs LM, Beaufrere H. Long term outcome after laser assisted modified Forssell's in cribbing horses. Veterinary Surgery, 2015 (Case series | Horse)DOI 10.1111/j.1532-950X.2014.12125.x
    Of the 90 horses followed up among the 119 operated on, 76, or 84.4 per cent, had not crib-bitten for over a year, 14 began again after six months at the median, crib-biting present for more than three years predicted failure, and 22.2 per cent had complications; welfare and colic risk were not recorded.