Risk Factors Associated with Colic in Horses PDF

Vet. Res. 33 (2002) 641–652  INRA, EDP Sciences, 2002 641 DOI: 10.1051/vetres:2002044 Review article Risk factors associated with...

Vet. Res. 33 (2002) 641–652  INRA, EDP Sciences, 2002 641 DOI: 10.1051/vetres:2002044 Review article Risk factors associated with colic in horses Stéphane GONÇALVESa*, Véronique JULLIANDa, Agnès LEBLONDb aLaboratoirede Nutrition des Herbivores, ENESAD, 26, boulevard du Dr Petitjean, BP 87999, 21079 Dijon Cedex, France bDépartement Hippique et Unité Epidémiologie et Modélisation des Maladies Transmissibles, École Nationale Vétérinaire de Lyon, 1, avenue Bourgelat, BP 83, 69280 Marcy l’Étoile, France (Received 29 November 2001; accepted 2 April 2002) Abstract – Many factors have been identified as risk factors for colic in horses in several epidemio- logical studies. The aim of our paper was to review the results of 12 epidemiological studies, in order to assess the impact of each risk factor for colic. According to the literature, the factors that increase the risk of colic are feeding practices (type and quality of food, type and changes of feeding), the in- trinsic factors of horses (sex, age and breed), management (type and changes of housing and activity), medical history (a previous colic, administration of a medical treatment) and parasite control (the presence of worms and type of deworming program). Several individual factors were incriminated as risk factors by all the studies. Nevertheless, the different studies did not always agree on the role of other risk factors. The conclusions were tightly related to several criteria in the selection of the study population, like the type of the epidemiological study, the number and the origin of horses included and the location of the study. colic / epidemiological studies / horse / risk factors / feeding practices Résumé – Facteurs de risque associés au syndrome colique chez le cheval. Différentes études épidémiologiques ont mis en évidence plusieurs facteurs de risque de coliques chez le cheval. L’ob- jectif de cet article a été de faire une synthèse des résultats de 12 études épidémiologiques, dans le but de déterminer l’impact de chaque facteur de risque dans l’apparition des coliques. D’après la littéra- ture, les facteurs augmentant les risques de coliques sont les pratiques d’alimentation (type et qualité des aliments, changements dans l’alimentation), les caractéristiques physiques des chevaux (âge, sexe, race), les pratiques d’élevage (type et changement de logement et d’activité), le passé sanitaire (historique des coliques, traitements médicaux) et le contrôle du parasitisme (présence de parasites, *Correspondence and reprints Tel.: (33) 3 80 77 29 24; fax: (33) 3 80 77 25 84; e-mail: [email protected] 642 S. Gonçalves et al. type de programme de vermifugation). Plusieurs facteurs de risque ont été identifiés par les 12 études épidémiologiques. Néanmoins, les différentes études ne sont pas toujours d’accord sur le rôle joué par certains de ces facteurs. Les conclusions de chaque étude semblent en effet étroitement liées à cer- tains critères dans la sélection de la population étudiée, comme le type d’étude réalisée, le nombre et l’origine des animaux inclus dans l’étude, le lieu de l’étude. coliques / étude épidémiologique / cheval / facteur de risque / pratiques d’alimentation Table of contents 1. Introduction............................................................. 643 2. Studies selected and methods used........................................... 643 3. Identified risk factors...................................................... 645 3.1. Feeding............................................................ 645 3.1.1. Type, composition and amount of food............................. 645 3.1.2. Distribution of food............................................ 646 3.1.3. Changes of feeding............................................. 647 3.1.4. Watering..................................................... 647 3.2. Internal parasitism.................................................... 647 3.2.1. Type of worms................................................ 647 3.2.2. Deworming treatments.......................................... 648 3.3. Intrinsic factors...................................................... 648 3.3.1. Sex......................................................... 648 3.3.2. Age......................................................... 648 3.3.3. Breed....................................................... 648 3.3.4. Other intrinsic factors........................................... 648 3.4. Medical history...................................................... 649 3.5. Management........................................................ 649 3.5.1. Housing..................................................... 649 3.5.2. Activity...................................................... 649 3.6. Weather-related factors................................................ 649 4. Comments on epidemiological methods and risk factors.......................... 650 4.1. Epidemiological methods used.......................................... 650 4.1.1. Number of horses included...................................... 650 4.1.2. Definition of colic............................................. 650 4.1.3. Origin of the horses............................................ 650 4.1.4. Origin of data and data collection................................. 651 4.1.5. Location of the studies.......................................... 651 4.2. Risk factors of colic.................................................. 651 5. Conclusion.............................................................. 651  Risk factors of horse colic 643 1. INTRODUCTION epidemiological studies were mainly used; they can be classified from those that are Digestive diseases, such as colic, diar- better at revealing risk factors to those that are rhoea, or enterotoxemia, represent 50% of less efficient: cohort studies, case-control the medical problems resulting in the death study designs and cross-sectional studies of adult horses. Colic, defined as ab- [2, 17, 31]: dominal pain of digestive origin, is the ma- (1) Cohort studies compare the inci- jor part of them: according to the National dence of one disease in an exposed popula- Animal Health Monitoring System tion and in a non-exposed population taken (NAHMS) in the USA, the incidence of as a control; subjects are included in the colic is estimated at 13.6% per year in a rep- study after their exposure has been deter- resentative sample of operations of horses mined and are followed forward in time to 6 months old or older in 28 states of the see if the disease appears. This type of study USA. Although 75% of the colic cases allows an accurate measurement of risk but were solved in less than 24 hours, 67% of generally requires a long period of time and the horses with colic were examined by a must include a large number of subjects. veterinarian, of which 85% received a treat- ment. The mortality rate for the horses The Cohort studies were the following: referred to veterinary clinics can be high Kaneene et al. observed 138 farms in [25, 34, 35]. Michigan for 2 years and collected data Every horseman has his own opinion of monthly on 3 925 horses. Tinker et al. [29, the cause of colic. Few scientific papers, 30] observed, for 1 year, colic events among however, have examined the aetiology of 1 427 horses from 31 farms in Virginia and colic and the ways to prevent it. Epide- Maryland. miological studies allow us to test the hy- (2) Case-control studies compare the pothesis of a relationship between exposure frequency of a previous exposure between to a risk factor and the development of a dis- subjects that develop a disease (cases) and ease. subjects that do not develop the disease of The aim of the present review was to interest (controls). The subjects are in- synthesise the information about different cluded in the study at the time of disease ap- factors that have been associated with pearance; then the exposure is measured. equine colic in published epidemiological This type of a study is less expensive than studies. This paper first presents the 12 epi- cohort studies since it is conducted over a demiological studies included and their ep- short period of time, with a limited number idemiological methods; the risk factors of cases and controls. The difficulty of this found in this literature are listed in the sec- type of a study is to well-determine the ond part; a third part comments on the va- characteristics of the controls, which must lidity and reproducibility of the results be comparable to the cases. obtained in these studies. The case-control studies were the fol- lowing: the first case-control study of colic used a teaching population at a hospital in 2. STUDIES SELECTED Minnesota that was examined over a period AND METHODS USED of 10 years. The study included 320 cases of colic and 3 610 controls and We identified 12 epidemiological stud- measured the association of colic with in ies concerning colic. Their main character- trinsic factors and medical management. To istics are shown in Table I. Three types of describe the different types of colic diagnosed  Table I. Characteristics of the 12 etiologic studies. 644 Reference Type of Period Number Number Number Horses treated Origin Data Country number study (Dates) of horses of Cases of Controls in clinic of data collect C.C.a 15 months 1 214 768 446 No Veterinarians Registration of USA (10/91–12/92) practice clinic and control horses C.C. 15 months 1 642 821 821 Yes Veterinarians Registration of USA (10/91–12/92) practice clinic and control horses C.C. 12 months 2 060 1 030 1 030 No Veterinarians Clinic files USA (03/97–02/98) practice C.a 2 years 3 925 77 / No Farms Registration of USA (02/92–01/93) (138 farms) practices health events (05/93–04/94) C.C. 2 years 300 200 100 No Veterinarians Registration of England (01/89–12/90) practice clinic and control horses C.S.a ? 231 116 115 No Hospital Feces England practice analysis S. Gonçalves et al. C.S. 18 months 206 103 sp. c.a 103 No Veterinarians Serums and England 60 20 il. i.a 40 practice feces analysis C.C. 10 years 3 930 320 3 610 Yes Hospital Clinic files USA (07/74–06/84) practice C.C. 11 months 812 406 406 Yes Hospitals Interview USA (03/91–01/92) practice by phone C. 1 year 1 427 151 / No Farms Events reported USA (11/90–11/91) (31 farms) practices on a calendar C. 1 year 948 86 / No Farms Events reported USA (11/90–11/91) (31 farms) practices on a calendar C.S. 5 years 4 644 / Yes Hospitals Clinic files USA – (1979–1984) practice England a C.: Cohort study; C.C.: Case/Control study; C.S.: Cross Sectional study; sp. c.: spasmodic colics; il. i.: ileal impactions.  Risk factors of horse colic 645 in England and associated risk factors, 3. IDENTIFIED RISK FACTORS 200 cases were compared to 100 controls from the practice of 7 English veterinarians Various risk factors associated with. Cohen et al. worked with 82 Texas colic were identified (Tab. II). These factors veterinarians, who collected data on one were classified into six categories: feeding; case of colic and one associated control parasitic infection; horse physical charac- monthly for 15 months. This report de- teristics; management; health history; cli- scribed 821 cases and 821 controls. A sec- matic conditions. ond paper concerned only the cases of chronic or intermittent colic. The same year, 5 clinics from the north-eastern USA 3.1. Feeding and Canada after a pilot study of 7 months, conducted a study lasting 11 months that 3.1.1. Type, composition and amount collected data from 406 cases of colic and of food 406 controls. Another case-control study conducted with 145 Texas veterinari- Horses are fed two main types of food: ans collected data from 2 060 horses forages and concentrates. Both have been (1 030 horses with colic and 1 030 non- incriminated as risk factors for colic, al- colic emergencies) to measure the associa- though no relationship between colic and tion between dietary and management fac- the type of feed was observed in Texas. tors and colic. * Forages: a high percentage of fibre has been suggested to favour the impaction of (3) Cross-sectional studies compare the the small intestine by stimulating intestinal frequency of an exposure between the cases motility, resulting in an important absorp- and controls. This type of study simulta- tion of water. In a study conducted on neously measures, at or over a defined mo- 18 horses involved in sporting events in ment of time, both disease and other Georgia and fed with a forage composed of variables of interest, in order to examine Cynodon spp., a graminae very rich in crude their relationships. It allows to have a tem- cellulose, 7 horses developed impaction. poral sequence of causes but the effect is The author assumed that a high percentage of not examined so it is not well-adapted for fibre in the diet increases the probability for etiological investigations. impaction. Hay of poorer quality and low digestibility has been shown to predispose The Cross-sectional studies included the horses to colic. However, other studies following: a first study aimed at the inven- have shown opposite results: a decrease in tory of the different types of colic, their di- colic was observed when horses only re- agnosis and the associated mortality rate in ceived forages [28, 30] or grass. 4 644 clinical cases from 14 American and 2 English universities. Proudman con- * Concentrates: while two studies con- ducted a coprologic analysis of 116 cases ducted in Texas did not show any relation and 115 controls from an English univer- between concentrate ingestion and colic sity to assess the association of tapeworms , the risk of colic increased with the (Anoplocaphala perfoliata) with colic. quantity of concentrate ingested ; feed- A subsequent study examined the associa- ing more than 2.5 kg of concentrate per day tion of tapeworms with either spasmodic appeared to increase the risk of colic by 4.8; colic or ileal impactions : 103 cases of more than 5 kg increased it by 6.3. spasmodic colic were compared to 103 con- Whole grain (except corn) was reported trols; 20 cases of ileal impactions were to reduce the risks of colic to 12% per in- compared to 40 controls. gested kilogram. Although barley and 646 S. Gonçalves et al. Table II. Risk factors cited in the 12 epidemiological studies. Risk factor Reference numbera Feeding practices Nature Forage X Concentrate X X Whole corn X X X Changes Changes of food X X X X X X X No regular watering X X X Parasitism Type Taenias X X Deworming Deworming treatments X X X No deworming program X X Intrinsic factors Sex Stallions X Geldings X Age Age from 2 to 8 years X X Age more than 11 years X X Breed Arabians X X X X X X X Thoroughbred horses X X Medical history Horses with previous colic X X X X X X Medical treatment X X X Management Housing Indoor stalling X Changes in housing X X X X X X X Activity Exercise X Intense activity X X X Stressing activity X Changes in activity X X X X Weather-related factors December, March and August X Weather change X a Cohen and Peloso, 1996. Cohen et al., 1995. Cohen et al., 1999. Kaneene et al., 1997. Proudman, 1991. Proudman and Edwards, 1993. Proudman et al., 1998. Reeves et al., 1989. Reeves et al., 1996. Tinker et al., 1997a. Tinker et al., 1997b. White and Lessard, 1986. oats were not associated with colic, the in- the amount of concentrate with colic [3, crease of the intake of concentrated food 15]. containing whole grain corn enhanced the Processed feeds, like pellets or risk by 3.4% per ingested kilogram in a sweetfeeds, have also been incriminated as case-control study, focused on 812 horses risk factors.. The arrival of concentrate in the colon causes a decrease in the luminal pH and 3.1.2. Distribution of food modifies the intestinal flora, favouring the production of endotoxins; these imbalances The mode and frequency of meals also could explain the observed association of have an impact on the apparition of colic: in  Risk factors of horse colic 647 pastures, horses spend 75% of the day and pears in the digestive system, transit is 50% of the night eating, whereas in boxes, modified and motility is altered. they are fed with a limited number of meals per day. The results of the epidemiological stud- ies regarding the role of parasites on the ap- parition of colic are controversial. On 3.1.3. Changes of feeding 3 925 horses, only 1.3% of the cases were A change in the quality or the quantity of reportedly associated with parasitism for food, as well as a change in the time of feed- 2 years (1 of 77 cases). ing or in the schedule of feeding result in an increased risk of colic [6, 7, 20, 26]. In a population of 1 642 horses (821 cases and 3.2.1. Type of worms 821 controls), the risk of colic was multi- plied by 2 during the 2-week period that fol- Strongyles (Strongylus spp.) have been lowed any change in feeding ; this risk described as the cause for colic in horses was 5 times greater in another case-control. These parasites could cause non stran- study conducted on 2 060 horses. A gulating colic, spasmodic in most of the study, conducted on 948 accumulated cases. The larvae were especially in- horse-years, showed that one change in criminated since they alter the motility of feeding during the year multiplies the risk the small intestine and they migrate in the factor for colic by 3.6. More than one blood vessels resulting in thrombolic disor- change during the year only multiplies the ders [18, 23]. risk factor by 2.2. Although effective anthelmintics have Among all the changes studied in feed- been used and in spite of the reduction of ing, that of the type of hay remains as the the prevalence of strongyles in equids, colic most significant factor ; the risk of colic cases due to parasites still occur. The was multiplied by 9.8 when the hay batch resistance of certain strongyle species changed in a study of 1 030 cases of colic (mainly Cyathostomum) to the anthelmintics and 1 030 controls. used, and the outbreak of “new” parasites like tapeworms (Anoplocephala perfoliata), 3.1.4. Watering which occupy the ecosystem, might explain the continuing problem. However, If not regular, watering increases the risk French et al. showed that deworming against of colic; a reduction in watering is caused strongyles does not increase the prevalence by several factors: the absence of a bucket of tapeworms. in the stall , a limited access to water, in paddocks or on dry lands , a too low wa- Before the 1980s, very few reports of ter temperature or a lack of water in cold tapeworm implication in equine intestinal weather [16, 33]. pathology were observed; since then, many clinical observations and epidemiological studies have shown that this parasite should 3.2. Internal parasitism not be neglected [14, 21, 22]. When the confounding factors (age, breed and sex of animals) are taken into account, the pres- Intestinal parasites, and especially tape- ence of tapeworms increases the risk of worms and strongyles (Strongylus vulgaris) ileal impaction by a factor of 3.4. Tape- are triggering factors for colic due to their worms are mass-located around the ileo- multiple actions: obstructive, traumatic, ir- caecal valve, which can disturb the transit ritating, toxic actions. As a result, pain ap- from the ileum to the caecum. 648 S. Gonçalves et al. 3.2.2. Deworming treatments of colic was observed between 2 and 8 years of age. In another study, the risk for In 821 horses treated by veterinarians colic was multiplied by 1.5 for horses older for colic and 821 control horses, no associa- than 10 years, compared to horses of 1 to tion between colic and the frequency of 5 years of age. Certain types of colic are deworming or the number of anthelmintics more frequent in certain categories of age used was detected. These results should [6, 20, 25]: colic resulting from the obstruc- be considered carefully since most of the tion of the intestine by enteroliths is fre- horses in this study were consistently fol- quent for horses over 11 years of age. lowed-up (daily anthelmintic treatment for Large colon impaction and displacement 30% of the horses, monthly for 26% and ev- seem to be more frequent in horses aged ery 2 months for 21%). In an ensuing study from 7 to 8 years whereas torsion and stran- in Texas, the risk for colic was 2.2 times gulation of the small intestine, large colon higher for horses which did not receive any and small colon obstruction decreases with regular deworming treatment. Uhlinger, age [4, 33, 34]. Moreover, colic treatments studying 3 groups of horses dewormed with requiring surgery are more frequent for different protocols, concluded that imple- older horses than for younger ones [4, 34]. menting a deworming program designed to But the relation between the age of horses have less than 200 eggs of strongyles per and the apparition of colic is complex. Age gram of feces, reduces the risk of colic. can be an indicator of the horses’ use, level A comparison between horses with colic of activity and feeding, which is related to and the control population established that the apparition of colic. an increase in the number of worming prod- ucts enhances the risk for colic. This result could be due to a biased selection: 3.3.3. Breed horses receiving better medical attention would be more easily treated for colic prob- Arabian horses seem to be particularly at lems. In another study, the risk for risk for developing colic [30, 34]: being an colic was multiplied by 2.1 in the 7 days fol- Arabian multiplies the risk for colic by lowing a deworming treatment. about 2 [6, 7]. The authors propose that management characteristics are responsible 3.3. Intrinsic factors for this prevalence of colic. Enterolithiasis is more frequent in Arabians or in miniature 3.3.1. Sex breeds. In another study, Thoroughbred Generally, no significant difference was horses were found to be the only breed at found in the distribution of colic according higher risk for colic. to sex [6, 7, 20, 25, 30, 34]. However, these results are still controversial: in a study 3.3.4. Other intrinsic factors conducted on 3 848 horses, the geldings ap- peared to be less affected than the stallions The size of the horse might also be an in- whereas, during the same year, the op- fluencing factor. In a study describing the posite result was shown in a study con- affected organs for the different types of ducted on 1 214 horses: geldings were colic, large warm-blooded breeds appear to more likely to colic. be more likely to be affected by dorsal dis- placement of the large colon: the intrinsic 3.3.2. Age factors of Hanovreian horses predispose them to a nephrosplenic entrapment. In a prospective study on 948 horses The impaction of the small colon appears to from 31 farms, an increase in the incidence be more frequent for ponies.  Risk factors of horse colic 649 Finally, the temperament, and especially with a change in diet and a change in the irritability and excitability of the horse activity [6, 7]. appear to be risk factors in a case-control study conducted on 800 equids. No association was found between colic and the type of bedding [6, 30]. However, the problem of straw as bedding is 3.4. Medical history well-known by horsemen: impaction colic often appears just after a change of bed- A horse is more likely to develop colic if ding, particularly if there is no other source it previously suffered from colic [6, 7, 26, of fibre. 33]. This risk is even greater if the horse has an abdominal surgery history: studies con- 3.5.2. Activity ducted on cases and controls showed that a horse with a surgical history has a risk No relation has been shown between the about five times greater than a horse with no activity of the horse and the occurrence of surgical history [6, 7]. However, the mortal- colic ; although an intense activity ity rate is not significantly different be- (training, competition) could increase the tween horses suffering from colic for the risk of colic [7, 23, 29]. Activities associ- first time and horses with a recurrent colic ated with stress have the same effects. history (respectively 12% and 17%). Moreover, an increased risk is associ- Medical treatment or vaccination in- ated with any change in activity [6, 7]; in a creases the risk for colic due to stress, espe- study comparing 1 030 cases to 1 030 con- cially if they have an effect on intestinal trols, the risk for colic was 2.2 higher for motility [16, 20]. On the contrary, no signif- horses which had a change in activity dur- icant association was shown between re- ing the 2-week period prior to examination cent vaccination and colic.. This result could also be explained by the correlated change in diet or in sta- In several studies, no relation was found bling. between dental disorders or the lack of den- tal care and colic, but many of these horses received regular dental care [6, 7, 30]. 3.6. Weather-related factors The association of colic with weather- 3.5. Management related factors has been considered by some investigators. A significant correlation was 3.5.1. Housing not found in measuring temperature varia- tions and barometric variations in the 24 h Housing conditions seem to influence preceding the colic crisis. No associa- the risk for colic: horses maintained on pas- tion was found between rainfall or tempera- tures all year long are less exposed to colic ture and the incidence of colic. than horses living indoors. Moreover, a relative risk factor of 1.6 was measured for Nevertheless, a study recorded more horses stalled more than 50% of the time colic cases during hot months ; another. It has been proposed that confinement, one noticed more cases during December, which generates stress and boredom, could March and August. A weather change increase the risk for chronic colic. during the 3-day period prior to examina- tion multiplied the risk of colic by 3.2. Changes in housing management partic- This result must be taken cautiously be- ularly predispose horses to colic [6, 7, 24, cause, as addressed by the author, the own- 30]. A change in housing is often associated ers of colic horses might have been more 650 S. Gonçalves et al. likely to remember weather conditions be- ther inclusion or exclusion parameters. fore the crisis than those of control horses. Proudman in 1991 made his analysis on six types of colic: spasmodic, flatulent, pelvic flexure impaction, other impactions, 4. COMMENTS surgical and colitis. In 1998, he included ON EPIDEMIOLOGICAL only two types: spasmodic colic and ileal METHODS AND RISK FACTORS impaction colic ; White and Lessard worked on obstructions, strangulation 4.1. Epidemiological methods used obstructions, peritonitis, non-strangulating infarctions and enteritis. Reeves et al. Establishing the association of factors excluded some diseases from their study related to colic in horses is not easy. Re- (enteritis or colitis, peritonitis, primary ported epidemiological studies point out ileus of unknown origin, broad ligament many factors that are responsible for colic, hematoma, undetermined surgical colic, re- but do not always agree on their respective current colic, uterine torsion and colic sec- role, bringing some confusion. However ondary to another primary problem). This their conclusions are tightly related to sev- lack of accuracy in the definition of colic eral criteria in the selection of the study probably led to a variability among the case population. selection between these 12 studies. This should be a potential source of bias in the 4.1.1. Number of horses included interpretation of the results; it can explain some of the differences observed in the im- In the 3 cohort studies, a great number of plication of the risk factors. Some risk fac- horses were observed for a long period of tors are actually only related to certain time (between one to five years). But the types of colic. number of horses with colic was not very high: 105 cases on average. On the contrary, a great number of cases were studied in 4.1.3. Origin of the horses case/control studies: in the 6 case/control studies in this review, the average number of cases was 590 horses and that of controls More than 30% of the epidemiological 1 068. Concerning the three cross sectional studies were only based on clinical records studies [21, 22, 34], the number of horses of the horses treated in veterinary clinics. with colic observed was respectively 116, This fact can introduce a bias in the selec- 123 and 4 644 cases. tion of cases and controls: only the most se- rious (and the less frequent) colic were 4.1.2. Definition of colic studied, especially when the study was con- ducted in a referral hospital; this does not In the 12 epidemiological studies, the allow to extend the results to every type of definition of colic varied: whereas colic. Moreover, in these studies, horses Proudman and Edwards bring together that were chosen as controls were often sick all the intestinal diseases to design colic, animals, so they did not accurately repre- the majority of the authors characterise this sent the healthy equine population. The syndrome with the notion of pain associ- three cohort studies [16, 29, 30] were run on ated to abdominal diseases [5–7, 29, 30]. the equine population of several farms. In Reeves et al. specify it as “abdominal this type of study a great number of healthy pain of digestive origin” and Kaneene et al. horses was selected and every type of colic relate it to distension of the stomach or was analysed, on the contrary to what was intestines. Other authors do not give a pre- studied with horses coming from veterinar- cise general definition of colic but list ei- ian clinics.  Risk factors of horse colic 651 4.1.4. Origin of data and data collection which received a medical treatment; in fact, these treatments often have some effects on 9 studies were based on veterinarian and the intestinal motility and microflora, hospital practices [5, 6, 7, 20–22, 25, 26, which might easily induce disorders of di- 34]; using this kind of data allowed a perti- gestion. Abrupt changes in activity, in nent collection of information concerning housing or in feeding practices were often the medical characteristics of the cases; but cited in the bibliography. But these factors information concerning the horse’s envi- were strongly correlated to each other so it ronment (feeding practices, activity, hous- is difficult to know exactly which factor can ing…) was generally poorer, especially induce a risk of colic. when studies were based on clinic files. On the contrary, three studies were based on farm practices [16, 29, 30]. Every event was 5. CONCLUSION reported on a calendar every-day, which provided much data and assured their accu- Many different factors have been associ- racy before their subsequent analysis as po- ated with the risk of colic in horses. Some of tential risk factors. them, like the intrinsic or weather-related ones, cannot be altered by human will. 4.1.5. Location of the studies However a better knowledge of their role should enable one to pay more attention. Half of the epidemiological studies were Most risk factors (feeding practices, para- run in different regions of the United States, sitism, management) depend on practices, and a third in England. The results were dif- and should be under human control when ficult to extrapolate to other regions or they are pointed out. More studies are countries, because management practices needed to precise the role of these practices. differed from one to another. Some studies For instance, the analysis of 12 epidemiologic grouped horses issued from American and studies showed that feeding practices appear English universities, which can introduce to be the most important risk factor for colic. bias in the results because these animals Dietary practices were cited as risk factors were bred and managed in different condi- in a fourth of the studies, however the asso- tions. ciation of nutrition and dietary factors with colic was specifically examined in only one 4.2. Risk factors of colic study. Even if everybody agrees that abrupt changes of food provokes colic, Table II lists the factors cited in the liter- other characteristics of feeding practices, ature as potential risk factors for colic and like the quality of food (and especially for- enumerates for each one the number of arti- age), the type and the rhythm of distribu- cles which identified it as a risk factor. This tion, the intake, etc., require more emphasises that all of these factors do not investigations to better understand their im- have the same importance for predisposing pact on this disease. to colic. Parasitism is the factor least cited in our REFERENCES bibliography. Intrinsic factors of horses (sex, age and breed) are often cited as po- Baker J., Ellis C., A survey of post mortem find- tential risk factors of colic because these ings in 480 horses from 1958 to 1980: (1) Causes factors are easily and routinely assessed. of death, Equine Vet. J. 13 (1981) 43-46. Concerning medical history, being a horse Bouyer J., Hémon D., Cordier S., Derriennic F., Stücker I., Stengel B., Clavel J., Epidémiologie : with previous colic appears to be an impor- principes et méthodes quantitatives, Les Edi- tant risk factor as well as being a horse tions INSERM, Paris, 1995. 652 S. Gonçalves et al. Clarke L.L., Roberts M.C., Argenzio R.A., Feed- cine and Surgery, ed. Médecine et Hygiène, 1997, ing and digestive problems in horses: Physiologic pp. 77-80. responses to a concentrated meal, Vet. Clin. Proudman C.J., A two years survey of equine North. Am., Equine Pract. 6 (1990) 433-450. colic in general practice, Equine Vet. J. 24 (1991) Cohen N.D., Epidemiology of colic, Surgical 90-93. Management of Colic: Vet. Clin. North Am., Proudman C.J., Edwards G.B., Are tapeworms Equine Pract. 13 (1997) 191-201. associated with equine colic? A case control Cohen N.D., Peloso J.G., Risk factors for history study, Equine Vet. J. 25 (1993) 224-226. of previous colic and for chronic, intermittent Proudman C.J., French N.P., Trees A.J., Tape- colic in a population of horses, J. Am. Vet. Med. worm infection is a significant risk factor for Assoc. 208 (1996) 697-703. spasmodic colic and ileal impaction colic in the Cohen N.D., Matejka P.L., Honnas C.M., Hooper horse, Equine Vet. J. 30 (1998) 194-199. R.N., Case-control study of the association be- Pugh D.G., Thompson J.T., Impactions Colics tween various management factors and develop- Attributed to decreased Water intake and Feeding ment of colic in horses, J. Am. Vet. Med. Assoc. Coastal Bermuda Grass hay in Boarding stable, 206 (1995) 667-673. Equine pract. 14 (1992) 9-14. Cohen N.D., Gibbs P.G., Woods A.M., Dietary Reeves M.J., What really causes colic in horses? and other management factors associated with Epidemiology’s role in elucidating the ultimate colic in horses, J. Am. Vet. Med. Assoc. 215 multi-factorial disease, Equine Vet. J. 29 (1996) (1999) 53-60. 413-414. Cohen N.D., Vontur C.A., Rakestraw P.C., Risk Reeves M.J., Gay J.M., Hilbert B.J., Morris R.S., factors for enterolithiasis among horses in Texas, Association of Age, Sex and Breed factors in J. Am. Vet. Med. Assoc. 216(2000) 1787-1794. Acute Equine Colic: a retrospective study of 320 Duncan J.L., Pirie H.M., The life cycle of cases admitted to a veterinary teaching hospital in Strongylus vulgaris in the horse, Res. Vet. Sci. 13 the USA, Prev. Vet. Med. 7 (1989) 149-160. (1972) 374-379. Reeves M.J., Salman M.D., Smith G., Risk fac- Foreman J.H., White N., Incidence of equine tors for equine acute abdominal disease (colic): colic in the University of Georgia ambulatory results from a multi-center case-control study, practice, Proceedings of the Second Equine Colic Prev. Vet. Med. 26 (1996) 285-301. Research Symposium, Athen, GA, 1986, p. 31. Rollins J.B., Clement T., Observations on Inci- French D.D., Chapman M.R., Klei T.R., Inci- dence of Equine Colic in a Private Practice, dence de 5 années de traitement à l’ivermectine Equine Pract. 1 (1959) 39-42. sur la prévalence du ténia chez le cheval, Vet. Rec. Tinker M.K., White N.A., Lessard P., Descriptive 135 (1994) 63-65. epidemiology and incidence of colic on horse Hassel D.M., Langer D.L., Snyder J.R., Drake farms: a prospective study, Proceedings of the 5th C.M., Goodell M.L., Wyle A., Evaluation of Equine Colic Research Symposium, Vol. 5, 1994, enterolithiasis in equids: 900 cases (1973– 1996), p. 22. J. Am. Vet. Med. Assoc. 214 (1999) 233-237. Tinker M.K., White N.A., Lessard P., Thatcher Holland J.L., Kronfeld D.S., Sklan D., Harris C.D., Pelzer K.D., Davis B., Carmel D.K., Pro- P.A., Calculation of faecal kinetics in horses fed spective study of equine colic incidence and mor- hay and concentrate, Am. Soc. Anim. Sci. (1998) tality, Equine Vet. J. 29 (1997) 448-453. 1937-1944. Tinker M.K., White N.A., Lessard P., Thatcher Jamme P., Les coliques d’origine parasitaire, Pra- C.D., Pelzer K.D., Davis B., Carmel D.K., Pro- tique Vétérinaire Equine spécial coliques 26 spective study of equine colic risk factors, Equine suppl. 4 (1994) 27-32. Vet. J. 29 (1997) 454-458. Julliand V., De Fombelle A., Drogoul C., Jacotot Toma B., Dufour B., Sanaa M., Benet J.J., Ellis P., E., Nutrition and microbial disorders in horses: Moutou F., Louza A., Epidémiologie appliquée à 3– Effects of three hay: grain ratios on microbial la lutte collective contre les maladie animales profile and activities, J. Eq. Vet. Sci. 21 (2001) transmissibles majeures, Maisons Alfort, 1996. 543-546. Uhlinger C., Effects of three anthelmintic sched- Kaneene J.B., Miller R.A., Ross W.A., Gallagher ules on the incidence of colic in horses, Equine R.K., Marteniuk J., Rook J., Risk factors for colic Vet. J. 22 (1990) 251-254. in the Michigan (USA) equine population, Prev. White N.A., Epidemiology and Etiology of Colic, Vet. Med. 30 (1997) 23-36. in: White N.A. (Ed.), Equine Acute Abdomen, Lilienfeld A.M., Lilienfeld D.E., Foundations of Malvern, USA, 1990, pp. 49-64. Epidemiology, 2nd ed., Oxford University Press, White N.A., Lessard P., Risk factors and clinical New York, Oxford, 1980. signs associated with cases of equine colic, Love S., The role of equine strongyles in the Equine Pract. Proc., 32nd Annual Conv. Am. As- pathogenesis of colic and the current options for sociation, 1986, pp. 637-644. prophylaxis, EquineVet. J., Suppl. 13(1992) 5-9. Wineland N., Baseline Reference of 1998 Equine Love S., Clinical aspects of equine intestinal para- Health and Management, National Animal sitism, in: 5th Geneva Congress of Equine Medi- Health Monitoring System, 1998.