Article Text

Download PDFPDF

Non-fatal horse related injuries treated in emergency departments in the United States, 2001–2003
  1. K E Thomas1,
  2. J L Annest1,
  3. J Gilchrist2,
  4. D M Bixby-Hammett3
  1. 1Office of Statistics and Programming, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, Atlanta, Georgia, USA
  2. 2Division of Unintentional Injury Prevention, National Center for Injury Prevention and Control
  3. 3American Medical Equestrian Association/Safe Riders Foundation, Albuquerque, New Mexico, USA
  1. Correspondence to:
 Karen E Thomas
 OSP/NCIPC/CDC, 4770 Buford Highway NE, Mailstop K-59, Atlanta, GA 30341, USA; KEThomas{at}


Objective: To characterise and provide nationally representative estimates of persons with non-fatal horse related injuries treated in American emergency departments.

Methods: The National Electronic Injury Surveillance System All Injury Program (NEISS–AIP) is a stratified probability sample comprising 66 hospitals. Data on injuries treated in these emergency departments are collected and reported. NEISS–AIP data on all types (horseback riding and otherwise) of non-fatal horse related injuries from 2001 to 2003 were analysed.

Results: An estimated 102 904 persons with non-fatal horse related injuries (35.7 per 100 000 population) were treated in American emergency departments each year from 2001 to 2003 inclusive. Non-fatal injury rates were higher for females (41.5 per 100 000) than for males (29.8 per 100 000). Most patients were injured while mounted on a horse (66.1%), commonly from falling or being thrown by the horse; while not mounted, injuries most often resulted from being kicked by the horse. The body parts most often injured were the head/neck region (23.2%), lower extremity (22.2%), and upper extremity (21.5%). The most common principal diagnoses were contusions/abrasions (31.4%) and fractures (25.2%). For each year that was studied, an estimated 11 502 people sustained traumatic brain injuries from horse related incidents. Overall, more than 11% of those injured were admitted to hospital.

Conclusions: Horse related injuries are a public health concern not just for riders but for anyone in close contact with horses. Prevention programmes should target horseback riders and horse caregivers to promote helmet use and educate participants about horse behaviour, proper handling of horses, and safe riding practices.

  • ATV, all-terrain vehicle
  • NEISS–AIP, National Electronic Injury Surveillance System All Injury Program
  • SR, sports and recreation
  • horse
  • equestrian
  • sports and recreation
  • traumatic brain injury

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Horseback riding is a popular recreational activity in the USA, particularly within certain demographic groups and regions of the country. In the USA, there are an estimated 9.2 million horses,1 with more than 19 million people aged 16 years and older participating in riding activities.2 Along with participation comes the risk of injury. A survey conducted of randomly selected households receiving a national horseback riding equipment catalogue found that 27.5% of participants aged 25 years and younger who rode six or more times a year had been treated by a physician in the past two years for a horse related injury, and 6.1% had been admitted to hospital for a horse related injury in their lifetime.3 Even though horse related activities have fewer participants than other sports and recreation activities, horseback riding is the eighth leading cause of emergency department treated, sports and recreation related injuries among female participants.4

Horseback riding and related activities have unique characteristics, with implications for the injuries sustained. Horse related sports are among the few sports in which participants from two different species function as a team, with the horse having the ability to act independently and unpredictably.5 In addition, horses are large and fast—often weighing 450 kg (1000 pounds) or more and travelling up to 48 km/h (30 mph)—with the rider’s head up to 3 m (over 9 ft) above the ground.5,6 Even when not mounted on a horse, a person can be seriously injured—a horse’s kick can generate a force up to 1.8 times its body weight.7

Although many studies describe injuries sustained in horse related incidents, most describe patients treated in a single medical centre.7,8,9,10,11,12,13,14,15 Most national studies conducted in the USA have focused on paediatric populations.16,17 This study provides current and comprehensive national estimates of all types (horseback riding and otherwise) of non-fatal horse related injuries treated in American hospital emergency departments, and characterises these injury incidents by demographics of the person injured, injury circumstances, types of injury, and disposition at discharge from the emergency department. In addition, previous injury prevention recommendations are discussed and placed in context using the surveillance data.


The National Electronic Injury Surveillance System (NEISS) is an emergency department surveillance system operated by the US Consumer Product Safety Commission.18 The system was designed to generate national estimates of consumer product related injuries. NEISS is a nationally representative stratified sample of 99 hospitals in the USA and its territories that have at least six beds and provide 24 hour emergency care. The hospitals are stratified into four categories based on the annual number of emergency department visits, with a fifth stratum for children’s hospitals.

The NEISS All Injury Program (NEISS–AIP) is a subsample of 66 of the 99 NEISS hospitals.19 In these emergency departments, data are collected on all injuries, regardless of cause or consumer product involvement. NEISS–AIP collects data on approximately 500 000 injury related emergency department visits annually, including age, sex, day and month of treatment, consumer products involved, primary body part injured, principal diagnosis, disposition at emergency department discharge, and a two line narrative describing the injuries and circumstances. Because the analysis of NEISS–AIP data does not involve patient contact, these data are exempt from institutional review board approval.

Cases were defined as patients injured while engaged in horse related activities (riding, mounting or dismounting, caring for a horse while not mounted, riding in a horse drawn buggy/cart, or being run over by a horse drawn vehicle). Cases were identified if one of two criteria was met: (1) if one of the two consumer product codes was “horseback riding—activity, apparel, or equipment”; or (2) if one of the product codes was “animal induced injury” and the narrative included the word “horse” or “pony.” Because NEISS–AIP does not capture all deaths, patients who were dead on arrival or who died while in the emergency department were excluded. Other types of cases were also excluded: motor vehicle occupants injured in a collision with a horse; injuries involving horse related equipment without mention of a horse (for example, tripped over a saddle on the floor); injuries associated with donkeys or mules; persons injured while in the barn or pasture with no mention of direct horse involvement (for example, falling while in the barn to feed horses); and persons who sought emergency department care because of a second unrelated injury incident (for example, motor vehicle crash today, fell off horse last week). In addition to the routinely collected NEISS–AIP variables, one author reviewed the narratives and coded additional circumstance variables. These included the mechanism of injury (fall, crush); whether the person was mounted, not mounted, or in the process of mounting/dismounting; other circumstances (horse tripped, person fell, and so on); and specific diagnoses (such as traumatic brain injury).

Each case was assigned a sample weight based on the inverse probability of selection and both non-response and post-stratification adjustments to account for changes in the total number of American hospital emergency department visits over time. These weights were summed to produce national estimates. The national estimates were based on weighted data for 4122 cases of horse related injuries seen in NEISS–AIP hospitals from 2001 to 2003 inclusive. Estimates were annualised by dividing the weighted totals by 3. Rates were calculated using the US Census 2001–2003 bridged race population estimates obtained from the National Center for Health Statistics.20 Confidence intervals were calculated using a direct variance estimation procedure that accounted for the sample weights and complex sampling design.


Each year from 2001 to 2003, an estimated 102 904 people (35.7 per 100 000 population) were treated in American hospital emergency departments for horse related injuries. About 19% of those were 14 years of age or younger (table 1). Overall, females had higher rates of injury than males, but sex specific rates and the differences between female and male rates varied across age groups (table 1, fig 1). Injury rates for females peaked at 10–14 years (90.6 per 100 000) and at 35–44 years (60.2 per 100 000). Ages 10–14 years also had the greatest difference in rates by sex, the female rate being more than three times the male rate. For patients aged 55 years and older, males had somewhat higher rates than females. Of all patients treated for horse related injuries, more than 11% were admitted to hospital or transferred for a higher level of care.

Table 1

 National estimates* and rates† of non-fatal horse related injuries treated in emergency departments, by age group, sex, and disposition at emergency department discharge—USA, 2001–2003

Figure 1

 National annual rates of non-fatal horse related injuries treated in emergency departments, by sex and age group—USA, 2001–2003.

Overall, the most common mechanism of injury was a fall (55.1%), followed by being struck by or against (24.0%) (table 2). Almost 6% of injuries were work related. The most common primary body parts injured were the head/neck (23.2%), lower extremities (22.2%), and upper extremities (21.5%). The most frequent principal diagnoses were contusions/abrasions (31.4%), fractures (25.2%), and strains/sprains (15.8%).

Table 2

 National estimates* of non-fatal horse related injuries treated in emergency departments, by selected injury characteristics, overall, and for those mounted or not mounted at time of injury—USA, 2001–2003

Nearly two thirds of injured persons were mounted on horseback at the time of the event causing the injury. For those mounted, the most common mechanism was a fall (78.9%), with the most common principal diagnoses being fractures (28.0%) and contusions/abrasions (27.9%). A majority (63.9%) of injuries were either to the trunk or head/neck region. For those not mounted, the most common mechanisms were being struck by/against (56.6%) and being crushed (25.6%), and the most common principal diagnoses were contusions/abrasions (40.8%) and fractures (18.9%). A majority (60.2%) of injuries for those not mounted involved the extremities.

Based on the principal diagnosis along with injuries described in the narrative, 11.2% of those injured had a traumatic brain injury (n = 11 502; 95% confidence interval (CI), 5663 to 17 340). The narratives also suggested that 1.6% of patients were treated for hip or pelvic fractures, and 1.5% sustained spinal fractures. Most injuries (55.7%) were treated in the months of May to September, and 40.1% were treated on Saturday or Sunday (results not in table).

The two most common circumstances of injury were first, falling or being thrown by the horse while mounted (44.6%), and second, being kicked by a horse while not mounted (11.2%) (table 3). Less common circumstances of injury included falling or being thrown while mounted because the horse reared, bucked, or was spooked (7.5%), and being stepped on by the horse while not mounted (6.1%).

Table 3

 National estimates* of non-fatal horse related injuries treated in emergency departments, by injury circumstances†—USA, 2001–2003

The distribution of injuries by primary body part affected and principal diagnosis differed for persons who were mounted and not mounted at the time of injury (table 4). Of those mounted, the most common diagnoses were traumatic brain injury (that is, concussions/internal injuries to the head) (11.9%); fractures/dislocations to the upper extremities (11.0%); fractures/dislocations to the upper trunk (9.0%); and contusions/abrasions to the lower trunk (8.0%). Of those not mounted, the most frequent diagnoses were contusions/abrasions to the lower extremities (18.2%), contusions/abrasions to the upper extremities (8.6%), fractures/dislocations to the lower extremities (7.5%), and lacerations to the head (7.2%).

Table 4

 National estimates* of non-fatal horse related injuries treated in emergency departments, by primary body part injured and principal diagnosis for those mounted or not mounted at time of injury—USA, 2001–2003


Most horse related injuries occur among females, which is in contrast to other injury causes which are often more prevalent for males.21 These injuries are quite severe, with a larger percentage admitted to hospital than for other causes of injury (11.2% v 5.5% for all injuries), and with a substantial number sustaining more serious diagnoses such as fractures and head injuries.21

The annual number of horse related injuries from 2001–2003 reported in this study (102 904) is higher than the estimates previously reported using NEISS. A study using 1990 NEISS data estimated that 74 349 persons with horse related injuries were treated in American hospital emergency departments.17 NEISS estimates for 2000 and 2001 (that is, 79 094 and 79 746, respectively) were reported in a newsletter for the American Medical Equestrian Association.22 Estimates in previous reports were based only on NEISS cases with a product code indicating “horseback riding—activity, apparel, or equipment,” whereas our estimates also used the narratives to identify other cases with horse related injuries. Our more comprehensive definition resulted in a 10-fold increase, from 2907 to 30 288, in the annual estimated number of cases identified as injured when not mounted.

The results observed in this study are generally consistent with previous studies. First, the greater number of females with horse related injuries in this study (59%) is consistent with the percentage of females seen in previous studies (52% to 85%).8,9,10,11,16,23 Second, the age patterns of those injured in this study are also consistent with previous studies, with most showing a peak for those aged 10 to 14 or 10 to 19 years,9,13,23 and some showing a second peak for those aged 30 to 40.13,23 Third, as in this study, most previous studies have found that falling/being thrown from the horse and being kicked are the two most common incidents leading to injuries.9–12,16,24,25 Finally, the figure of 11% of patients admitted to hospital or transferred in this study corresponds to hospital admission rates in other studies (7% to 30%).9,10,13,14,23,26–28

Horse related injuries share some similarities and dissimilarities with other sports and recreation (SR) injuries. Both horse related and SR injuries peak among adolescents aged 10 to 14 years4; however, SR injuries generally do not peak a second time in the 35 to 49 year age group. Although the SR injury rate for males is twice that for females, the horse related injury rate is 1.4 times higher for females than for males. Horse related injuries are also more severe, as evidenced by 11.2% being admitted to hospital or transferred, as opposed to 2.3% for SR injuries.

Comparisons between horse related injuries and all-terrain vehicle (ATV) injuries are relevant. Both involve potentially high speeds, an “off road” environment, and riders external to both “vehicles.” Each year, about 127 000 ATV related injuries are treated in American hospital emergency departments, similar to the number of horse related injuries, and a comparable percentage is admitted to hospital or transferred (12.2% for ATV injuries v 11.2% for horse related injuries). The composition of serious injuries as determined by principal diagnoses is also similar, with 26.0% of ATV related patients sustaining fractures and 7.8% diagnosed as having traumatic brain injuries, compared with 25.2% and 9.7%, respectively, for patients with horse related injuries (unpublished NEISS-AIP data). Given the similarity of horse related and ATV injuries, surprisingly safety regulation/legislation only involves ATV riders (helmet use, minimum age requirements, and safety education)29; no such regulations exist for horseback riders.


This study had several limitations. First, because data are lacking on the number of participants in horse related activities, injury rates were calculated based on US population estimates and could not account for different exposures to horses by age and sex. Second, because the narratives varied in detail, data were not systematically available on factors potentially associated with the injury incident, such as protective equipment use (helmets, vests); alcohol use; skill level of the rider; information about the horse’s nature or behaviour; and environmental conditions. Third, NEISS–AIP data only capture injuries treated in hospital emergency departments, not injuries treated elsewhere or not at all. Fourth, NEISS–AIP data only allow one body part diagnosis to be recorded; when possible, however, other severe injuries mentioned in the narratives were coded. Finally, data are not presented by race because of the high percentage (26%) of cases with no race specified.

Implications for prevention

The frequency and severity of horse related injuries indicate that prevention programmes are needed. Measures can be taken to prevent horse related injuries, either by reducing the likelihood of incidents that may result in injuries (primary prevention: securing saddle, education) or by reducing the likelihood of injuries once an incident has occurred (secondary prevention: helmet use).

Probably the most important measure individuals can take to reduce serious injuries is to wear a helmet meeting the standards of the American Society for Testing and Materials (ASTM)/Safety Equipment Institute (SEI), such as ASTM standard F1163, or a similarly certified helmet. Currently, the American Academy of Pediatrics, the American Medical Association, the US Pony Club, and other organised groups recommend the use of ASTM/SEI helmets.5,6,30 The mechanisms for head injury are similar to those for activities in which helmets have proven effective (riding a bicycle, motorcycle, or ATV), and many horse related studies have shown either a reduction in injuries following increased helmet use or less severe injuries among people wearing helmets.5,7,15,31 Helmets should fit well and be properly secured.5 Because of the risk of being kicked while not mounted, some experts have recommended wearing helmets even when not mounted for young children and for certain activities (Malavase D, personal communication); in this study, 16% of the patients with a principal diagnosis of traumatic brain injury sustained the injury while not mounted.

Unfortunately, there are no current estimates of equestrian helmet use in the USA. The latest survey conducted in 1991 estimated that 43% of riders wore helmets at least some of the time.31 Identified barriers to helmet use (for example, low risk perception, lack of comfort and style) apply to both equestrian and bicycle helmets31,32; promotion strategies identified for bicycle helmets may also be successful in promoting helmets for horse related activities.33,34

Previous recommendations emphasise wearing appropriate riding attire. Sturdy boots with a heel should be worn to prevent the feet from slipping through the stirrup.3,5,6,24,27,35–38 Gloves of non-skid material may also be worn.5,35,39 Loose fitting clothes should be avoided and hair should be tied back to avoid entanglement with branches and other obstacles.3,6,27,36

Other safety equipment has been suggested, although the protective effects have not been proven. Safety stirrups that release when a certain amount of pressure is applied may prevent a rider from being dragged. Though recommended by some experts,9,37 others question whether they are effective.5,35,36 In this study, more than 350 people were injured each year when a foot was caught in a stirrup. Additionally, body protector vests may protect the spinous processes and ribs from kicks and falls.36 While mandatory for certain events in Britain,25 the effectiveness of such vests has not been evaluated.5 An ASTM standard for these vests has been established (ASTM F1937).

Individuals should be in the proper physical and mental condition to ride or work around horses.35 Riding requires balance, agility, mental acuity, and a reasonable level of physical fitness.5,40 Alcohol should not be consumed when riding or working with horses.40 Previous studies have found that of those tested, 17% of horse related hospital admissions12 and 33% of deaths39 involved a detectible blood alcohol level. In this study, less than 1% of unweighted cases (34) were reported to involve alcohol use. However, alcohol use was probably underreported because the information was not routinely collected in NEISS-AIP.

Education by trained instructors may help prevent horse related injuries.5,6,17,24,27,30,40 Horses are a prey species whose actions can be better understood and predicted when the animal’s instincts are considered. For example, horses can be spooked when exposed to new experiences or perceived threats (for example, a person walking into a horse’s blind spot).38 Additionally, experts recommend instruction in proper falling techniques.5,24,27,36,40

Aside from protective equipment, appropriate clothing, and training, other safety measures should be considered when preparing to ride. One should ensure that the horse is in good health and appropriately matched to the rider’s abilities.3,5,30,37,40 Additionally, inspect all tack to ensure that it is correctly applied.3,5,6,27,40 In this study, an estimated 550 people were injured each year when the saddle slipped or broke. Riders and others working with horses should avoid wrapping the reins, lead, or rope around their hands.37,38 An estimated 2000 people in this study were injured annually when their hands or fingers were caught in the horse’s tack.

Riding organisations and stables can influence proper horse riding and handling by instituting regulations or guidelines regarding helmet use, appropriate attire, rider education, and supervision to create an atmosphere where safe practices are routine. Previous researchers suggest that children and inexperienced riders should be adequately supervised,5,37 and children under six should not be allowed to ride independently.6 In addition, riding environments can be modified to reduce the risk of injury (for example, energy absorbing riding surfaces5). Finally, previous researchers recommend having medical care available at organised events,5,41,42 along with having an emergency plan for daily operations including first aid training for personnel and the availability of emergency supplies.43

Riders, equestrian organisations, and health care providers have opportunities to prevent horse related injuries by counselling those involved in these activities. Stressing injury prevention to persons who have experienced an injury may be beneficial because reinjury rates are reportedly high, with 25% to 37% of injured persons having had a previous injury.9,10,23,28 Because of the frequency of concussions during horse related activities and the concern regarding repeated concussions, the use of return-to-play guidelines similar to those used in contact sports is relevant.24,44

What is already known on this topic

  • Previous studies suggest that horse-related injuries can be severe and are more frequent in females, particularly those ages 10-14 years

What this study adds

  • Over 100 000 people with horse related injuries are treated in American emergency departments annually. Using a national sample, this study characterises these injuries overall, and when mounted or not mounted

  • Horse related injuries can be severe as evidenced by higher percentages of hospital admissions, fractures, and traumatic brain injuries when compared with other recreation related injuries


This study shows that horse related injuries are a serious issue, particularly for young females. The hospital admission rate and the percentage of people with potentially severe diagnoses underscores the need for prevention efforts. Prevention programmes should target helmet use as well as education and training about horse behaviour, and safe horse handling and riding. Additional research is needed to evaluate the effectiveness of specific interventions to reduce the number and severity of horse related injuries.


We thank Tom Schroeder, MS, Director, Cathy Irish, and other staff members of the Division of Hazard and Injury Data Systems, US Consumer Product Safety Commission, for collecting and preparing data in collaboration with the Office of Statistics and Programming, National Center for Injury Prevention and Control (NCIPC). We also acknowledge Patricia Holmgreen, formerly of NCIPC, for her contributions to the data preparation process, and Drusilla Malavase for her insightful comments on the manuscript.



  • Published Online First 12 April 2006

  • Competing interests: none declared