Fencing injury research, one study at a time.

The studies side by side

Each row counts something different, so the figures cannot be compared row against row.

The IOC chapter: a review of earlier studies (Harmer, 2010)

Peter Harmer's chapter on fencing in the IOC Medical Commission's Epidemiology of Injury in Olympic Sports (2010), which the FIE hosts with IOC permission, reviews the studies published up to then. Its main findings, each from the study it names:

• Studies of time-loss injuries consistently found 0.0 to 0.3 per 1,000 athlete exposures across competition and training settings. By contrast, 20% to 27.5% of fencers were recorded as injured at particular competitions, and 59% to 77.8% reported an injury at some point in their careers. The chapter reads the exposure-based rates as showing most fencing injuries are minor.
• Most studies found the legs the most common site (mean 59%, range 42.9% to 77.3%), especially the ankle and knee; several instead found the wrist, hand and fingers most often injured (54.6% to 61.3%).
• Gambaretti and colleagues (1992; 178 Italian fencers over one year) recorded 77% of injuries in practice and 23% in competition. The chapter notes it is unclear whether the rate per exposure is higher in practice.
• Age: Harmer (2008b, a conference abstract covering about 42,000 participants) found no significant difference in time-loss rate between youth fencers aged 8 to 17 (0.27 per 1,000 exposures), young adults aged 18 to 30 (0.31) and veterans aged 50 and over (0.21).
• Weapon, Italian data: Majorano and Cesario (1991), using the 1991 Italian national competition, noted about twice as many foil fencers injured (18.6%) as epee (8.9%) or sabre (9.2%) fencers, but did not test the difference for significance.
• The chapter names three main causes in the literature: hits from the opponent's weapon, poor technique, and the piste itself. Slipping, tripping or stepping off the piste edge accounted for between 9.6% (Carter and colleagues, 1993) and 37% (Gambaretti and colleagues, 1992) of injuries in those studies.
• It states that no fencing injury prevention studies had been done, and that FIE safety-related equipment and rule changes had been based on face validity, with no research testing whether any of them worked.

US national competitions (Harmer, 2008a)

Harmer's prospective five-year study in the Clinical Journal of Sport Medicine covered about 78,223 competitors at US national competitions (47,483 male, 30,740 female). The figures here are as the IOC chapter tabulates them; the original paper was not read.

• 184 time-loss injuries: 0.3 per 1,000 exposures overall, 0.27 for men and 0.36 for women.
• The knee was the most often injured region (19.6%), and ACL injuries were singled out as a particular concern.
• Injury types: strain 26.1%, sprain 25.5%, contusion 12.0%, subluxation or dislocation 7.6%, fracture 4.4%, puncture 2.7%, laceration 0.5%.
• Women had a 35% higher risk of a time-loss injury than men (relative risk 1.35).
• Foil and epee did not differ significantly. Sabre carried a 62% higher risk than foil and epee (relative risk 1.62), seen in both men (1.61) and women (1.63).

FIE international competitions (Harmer, 2019): unconfirmed

Harmer's five-year analysis of FIE competitions was published in the British Journal of Sports Medicine (2019; citation confirmed on the IAT sport literature database). The journal's pages returned an error when the research tried to open them, so every figure below comes from a search excerpt of the abstract and is unconfirmed.

• 809 FIE competitions in the 2010 to 2014 seasons; 176 time-loss injuries over 637,776 exposures, or 0.28 per 1,000.
• Men had a higher risk than women (relative risk 1.42).
• Epee had a lower risk than foil (relative risk 0.52) and than sabre (0.47).
• Sprains (40.8%) and strains (20.1%) were the majority; 72.4% were in the legs; ankle sprains were the most common single injury (25.3%); non-contact effort by the injured fencer was the leading mechanism (47.1%). Median time lost was 4 weeks, and 32.1% of injuries meant 2 weeks or less away.

French elite fencers at INSEP (Gondouin et al., 2025)

This retrospective study covered all elite fencers at INSEP, France's national sport institute, from June 2016 to May 2023: 117 fencers (56 female). Sports medicine physicians recorded every musculoskeletal complaint or concussion, whether or not time was lost. Training hours were the study's own estimate (1,056 per fencer per year).

The study found no significant difference between weapons, or between sexes within a weapon. Overall, 71% of injuries were in the legs (60.6% of those in the front leg), 18.4% in the arms (96.3% of those in the sword arm) and 10.4% in the trunk; the thigh (20.3%), knee (18.8%) and ankle (11.2%) were the most injured sites. Epee fencers had more knee and forearm injuries than foil and sabre fencers; foil fencers had fewer foot and ligament injuries. Women had more pelvis and hip injuries than men (10% against 5.3%).

The 2025 meta-analysis (Grivas et al.)

Grivas and colleagues reviewed 19 studies published from 1990 to 2022, covering 1,151 able-bodied fencers in prevalence studies, and 164,485 able-bodied and 14 wheelchair fencers in incidence studies. The findings here are from the summary published by the authors' university, The Hong Kong Polytechnic University.

• Prevalence among able-bodied fencers ranged from 51.5% to 92.8% across studies; not all findings could be pooled.
• Weapon was not consistently linked to injury among non-professional fencers. Low-quality evidence suggested professional sabre fencers may have a higher time-loss risk than epee fencers. No finding was reported for foil.
• Leg injuries were more than half of those reported in able-bodied fencers; in wheelchair fencers, shoulder, elbow and forearm injuries were most common.
• Older age and longer experience were linked to more injury among non-professional fencers. The authors concluded that fencing injuries are common, especially among professional and experienced athletes.

US emergency-department visits (Stanicki et al., 2025)

This study used NEISS, the US Consumer Product Safety Commission's emergency-department sample, for 2013 to 2023. NEISS counts hospital visits, not fencers, so its figures describe who came to an emergency department and with what, never the risk to any fencer.

• 129 cases, a weighted national estimate of 3,418 visits. Median age 16 (range 4 to 68); 35.7% were aged 11 to 15 and 29.5% aged 16 to 20; 56.5% were male.
• Diagnoses: strain or sprain 26.4%, laceration 15.5%, fracture 11.6%, contusion or abrasion 10.9%, concussion 4.7%, puncture 3.9%, dislocation 3.9%, internal organ injury 1.6%. Fractures were more common in female patients (14.4% against 9.6%), lacerations in male patients (21.9% against 7.1%).
• The finger (14%), wrist (8.5%), leg (8.5%) and head (7.8%) were the most affected parts. 97.7% were treated and released.
• Weapon was recorded in only about 5% of cases (2 foil, 5 sabre, no epee), so the data cannot compare weapons.

Wheelchair fencing

No wheelchair-fencing injury rate was found. Willick and colleagues' study of the London 2012 Paralympic Games gives all-sport figures (12.7 injuries per 1,000 athlete-days), not fencing figures, and its sport-by-sport table could not be opened. A 2015 German review (Hirschmueller and colleagues) lists wheelchair fencing among the Summer Paralympic sports with the highest injury rates, without a fencing figure.

Where the studies disagree

• Sex. Harmer 2008a (US national competitions) found women at higher time-loss risk; Harmer 2019 (FIE competitions, unconfirmed) reported men at higher risk; the INSEP study found no significant difference; the 2025 meta-analysis found no clear overall difference among professional fencers.
• Weapon. Harmer 2008a found sabre at higher risk than foil and epee; Harmer 2019 (unconfirmed) reported epee at lower risk than foil and sabre; INSEP found epee numerically highest but no significant difference; the Italian 1991 data showed more foil fencers injured, untested.

The populations, definitions and measures differ, so none of these figures is combined with another. Puncture wounds and deaths are covered on penetrating injuries and deaths.