| Study | Sport | Study Design | Duration (seasons) | Study Population | Exposure Measures | Outcome Measures | Results |
| Wisniewski (2004)26 | Football | PC | 1 | 87 NCAA Division I-A teams | Custom vs non-custom MG use | Concussion | No significant difference in incidence of concussions between players wearing custom vs non-custom MGs (95% CI: 0.99 to 1.75) |
| McNutt (1989)32 | Football | XS | 3 | 21 high school teams (n = 2470) | MG use | Sport-related injuries | 56% of all concussions sustained while not wearing MG (p = 0.001) |
| Garon (1986)35 | Football | XS | n/a | 20 high school teams (n = 754) | MG use | Concussion | 48% of concussions sustained while not wearing MGs |
| Stenger (1964)13 | Football | XS | 5 | 1 College team | Custom MG use | Head and neck injuries | 90% of concussions sustained during 1 season were by players who did not wear MGs |
| Anonymous (1972)38 | Football | CS | 1 | 1 University team | Custom MG use | Concussion | No concussions for MG users vs 2 concussions for non-users |
| Dorney (1994)36 | Rugby | CR | 1 | First grade player (n = 1) | MG use (triple-laminated) | Concussion and Mandibular Fracture | No recurrent concussion symptoms while using triple-laminated pressure moulded mouthguard |
| Blignaut (1987)34 | Rugby | XS | 1 | First team players (n = 321) | MG use | Head and neck injuries | No statistically significant differences in concussion rates between wearers and non-wearers of MGs |
| De Wet (1981)33 | Rugby | XS | 1 | 10 primary schools (10–13 year olds) (n = 150) | Custom MG use | Concussion | No concussions for MG users, 12% concussions for non-users |
| Marshall (2005)40 | Rugby | PC | 1 | Rugby Union players in Dunedin, New Zealand n = 304 | Self-report use of MGs | Concussion | No evidence of a protective effect of MGs (RR = 1.62, 95% CI: 0.51 to 5.11) |
| Barbic (2005)42 | Football & rugby | CRCT | 1 | 5 Ontario Universities: male football (n = 394), male rugby (n = 129), female rugby (n = 123) | WIPSS Brain-Pad MG vs all other MGs | Concussion | No significant difference in number of concussions between WIPSS MG use and all other MGs (p = 0.79; OR: 1.06 (95% CI: 0.51 to 1.61) |
| Finch (2005)43 | Australian rules football | CRCT | 1 | 23 teams (n = 301) | Custom MG vs usual MG behaviours | Head/orofacial injuries | Concussions were distributed roughly equally across study arms (underpowered to be able to assess statistically) |
| Labella (2002)19 | Basketball | PC | 1 | 50 NCAA Division I teams | MG use | Concussion | No significant differences in concussion rates (0.35 vs 0.55) or oral soft tissue injury rates (0.69 vs 1.06) between MG users and nonusers, respectively |
| Benson (2002)39 | Hockey | PC | 1 | 22 CIAU teams (n = 642) | Full vs half face shield use, MG use | Concussion severity measured by time loss due to injury | Half shield cohort: MG use: 2.76 sessions lost per concussion (95% CI: 2.14 to 3.55, n = 23 concussions), no MG use: 5.57 sessions lost per concussion (95% CI: 4.40 to 6.95, n = 14 concussions) |
| Full shield cohort: MG use: 0 time loss (no concussions), no MG use: 1.80 sessions lost per concussion (95% CI: 1.38 to 2.34, n = 2 concussions) | |||||||
| Benson (2005)41 | Hockey | PC | 1 | National Hockey League (n = 1033) | MG vs no MG | Concussion and concussion severity | The risk of concussion for athletes who did not wear a MG was 1.42 times greater than players who wore a MG, but this difference was not statistically significant (95% CI 0.90 to 2.25) |
| Concussion severity, measured by time loss from competition, was not significantly different between the two cohorts | |||||||
| Symptom severity measured subjectively using the modified McGill ACE symptom scale was significantly greater for athletes not wearing MGs compared with those who did (p<0.01) | |||||||
| Mihalik (2007)37 | Any | CS | n/a | Athletes that sustained a sports-related cerebral concussion (n = 180) | Self report use of MGs | Neurocognitive impairments using ImPACT | No difference in neurocognitive deficits at time of first follow-up assessment (mean 3 days) between MG users and non-users |
| Takeda (2005)44 | n/a | Lab | n/a | Artificial skull model | MG vs no MG | Surface distortions related to bone and acceleration of the head | MG use significantly decreased distortion of the mandibular bone and acceleration of the headform compared with no MG (p<0.01) |
| Hickey (1967)12 | n/a | Lab | n/a | 1 male cadaver | 2 types of MG | Pressure changes and bone deformation within the skull | 50% reduction in bone deformation and intracranial pressure amplitude with a MG in place |
CR, case report; CS, case series; CRCT, cluster-randomised controlled trial; CIAU, Canadian Inter-University Athletics Union; NCAA, National Collegiate Athletic Association; MG, mouthguard; PC, prospective cohort; XS, cross-sectional