Abstract
This study uses a national survey of US adolescents to examine the prevalence and correlates of self-reported diagnosed concussions to target and monitor prevention efforts.
Little is known about the prevalence and correlates of concussions among US adolescents. A regional study of Canadian adolescents found that approximately 20% had sustained a concussion. Providing a national baseline of concussion prevalence and correlates is necessary to target and monitor prevention efforts to reduce these types of injuries during this important developmental period.
Methods
This study used cross-sectional data from the 2016 Monitoring the Future (MTF) survey, an annual, in-school survey of US students in grades 8, 10, and 12. Response rates for the 2016 MTF survey were 90% for grade 8, 89% for grade 10, and 81% for grade 12. The institutional review board at the University of Michigan approved this study. A waiver of informed consent was sent to parents providing them a means to decline their child’s participation if necessary.
An item to assess concussion was developed in collaboration with concussion experts, pilot tested, and added to the MTF survey in 2016. Respondents were asked: “Have you ever had a head injury that was diagnosed as a concussion?” Response options included “no,” “yes, once,” and “yes, more than once.”
Sociodemographic variables included sex, race/ethnicity, grade level, and participation in competitive sport within the past 12 months. Participation in competitive sport assessed 21 different types of sports and was recoded into a mutually exclusive variable that included 5 participation categories: no sports, contact sports, semicontact sports, noncontact sports, and other sports (see Table 1 for details).
Table 1. Characteristics of US Adolescents Reporting a Diagnosed Concussion in 2016 (N = 13 088)a.
| No. of Adolescents (%) [95% CI] | |
|---|---|
| Sex | |
| Male | 6177 (49.8) [48.6-51.0] |
| Female | 6223 (50.2) [48.9-51.4] |
| Race/ethnicityb | |
| White | 5877 (46.8) [43.4-50.1] |
| Black | 1600 (13.1) [10.8-15.3] |
| Hispanic | 2674 (19.1) [16.4-21.9] |
| Otherc | 2837 (21.0) [19.3-26.7] |
| Grade leveld | |
| 8 | 5885 (32.6) [27.0-38.2] |
| 10 | 5100 (33.9) [27.8-39.9] |
| 12 | 2103 (33.5) [27.3-39.6] |
| Sport participation in past 12 mo | |
| No sports | 3040 (26.3) [24.9-27.6] |
| Contact sportse | 2851 (21.8) [20.6-23.1] |
| Semicontact sportsf | 3596 (28.7) [27.3-29.9] |
| Noncontact sportsg | 2085 (16.8) [15.7-17.8] |
| Other sportsh | 768 (6.4) [5.7-7.1] |
| Diagnosed concussions in lifetime | |
| 0 | 10 175 (80.5) [79.5-81.6] |
| 1 | 1709 (14.0) [13.1-14.8] |
| >1 | 633 (5.5) [4.9-6.1] |
Of respondents, 4.4% had missing data for number of diagnosed concussions and 5.7% for sport participation. Multiple imputation was used to impute missing observations. Several separate imputations with the 2016 Monitoring the Future survey sample used chained multinomial and logistic in STATA’s “mi impute chained” procedure (10 imputations, 5 burn-in iterations each). Percentages and 95% CIs were based on weighted data; samples sizes were based on unweighted data.
Measures for race/ethnicity were self-reported and predefined within the survey. Race was assessed because of differences across behavioral and social outcomes.
Other race includes Asian, American Indian, those who selected multiple races/ethnicities, and those with missing racial information.
Weights were used to adjust for the different sample sizes for 8th, 10th, and 12th graders.
Contact sports included football, ice hockey, lacrosse, and wrestling. Involvement in any of these sports (regardless of any additional sports) was defined as participating in contact sports.
Semicontact sports included baseball, basketball, field hockey, and soccer. Involvement in any of these sports (regardless of additional noncontact and “other” sports) was defined as participating in semicontact sports.
Noncontact sports included cheerleading, crew, cross-country, equestrian, golf, gymnastics, swimming, tennis, track, volleyball, water polo, and weightlifting. Involvement in any of these sports (regardless of additional “other” sports) was defined as participating in noncontact sports.
Other sports indicates nonspecified other types of sport. Involvement in any of these sports was defined as participating in other sports.
Binary logistic regression models (using STATA [StataCorp], version 14.0) were used to estimate odds ratios (ORs), adjusted odds ratios (AORs) and 95% CIs. All binary logistic regression models estimating the AORs simultaneously controlled for sociodemographic variables (listed in Table 2). Analyses used weights to account for the probability of selection into the sample and adjust for the different sample sizes for grades 8, 10, and 12. Two-tailed significance tests with an α of .05 were used to assess statistical significance. Given that 10.4% of the sample had missing data on at least 1 of the variables used in the analyses, multiple regression imputation was used to impute missing data (see Table 1 for details).
Table 2. Correlates of Lifetime Prevalence of Diagnosed Concussion Among US Adolescents in 2016 (N = 13 088).
| Any Diagnosed Concussion | 1 Diagnosed Concussion | >1 Diagnosed Concussion | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Adolescents, % | OR (95% CI) | AOR (95% CI)a | Adolescents, % | OR (95% CI) | AOR (95% CI)a | Adolescents, % | OR (95% CI) | AOR (95% CI)a | |
| Sex | |||||||||
| Male | 23.2 | 1 [Reference] | 1 [Reference] | 16.0 | 1 [Reference] | 1 [Reference] | 7.3 | 1 [Reference] | 1 [Reference] |
| Female | 15.8 | 0.62 (0.55-0.68) | 0.83 (0.72-0.95) | 12.1 | 0.73 (0.63-0.85) | 0.88 (0.77-1.02) | 3.7 | 0.48 (0.38-0.62) | 0.72 (0.54-0.94) |
| Race/ethnicity | |||||||||
| White | 23.7 | 1 [Reference] | 1 [Reference] | 16.5 | 1 [Reference] | 1 [Reference] | 7.3 | 1 [Reference] | 1 [Reference] |
| Black | 14.7 | 0.55 (0.45-0.66) | 0.52 (0.42-0.64) | 12.2 | 0.72 (0.57-0.88) | 0.68 (0.55-0.85) | 2.5 | 0.32 (0.20-0.52) | 0.29 (0.17-0.48) |
| Hispanic | 13.1 | 0.48 (0.41-0.56) | 0.51 (0.44-0.61) | 10.2 | 0.57 (0.48-0.69) | 0.61 (0.50-0.73) | 3.0 | 0.39 (0.25-0.60) | 0.42 (0.27-0.64) |
| Otherb | 18.8 | 0.74 (0.64-0.87) | 0.77 (0.66-0.89) | 13.3 | 0.78 (0.65-0.94) | 0.82 (0.68-0.97) | 5.5 | 0.72 (0.54-0.95) | 0.78 (0.58-1.05) |
| Grade level | |||||||||
| 8 | 17.2 | 1 [Reference] | 1 [Reference] | 13.2 | 1 [Reference] | 1 [Reference] | 4.0 | 1 [Reference] | 1 [Reference] |
| 10 | 19.4 | 1.16 (1.01-1.33) | 1.24 (1.09-1.42) | 13.5 | 1.02 (0.89-1.17) | 1.07 (0.93-1.22) | 6.0 | 1.54 (1.19-2.01) | 1.68 (1.32-2.16) |
| 12 | 21.8 | 1.35 (1.14-1.61) | 1.49 (1.27-1.75) | 15.4 | 1.19 (1.01-1.41) | 1.29 (1.08-1.54) | 6.5 | 1.64 (1.21-2.21) | 1.88 (1.38-2.56) |
| Sport participation in past 12 moc | |||||||||
| No sports | 11.4 | 1 [Reference] | 1 [Reference] | 9.1 | 1 [Reference] | 1 [Reference] | 2.3 | 1 [Reference] | 1 [Reference] |
| Contact sports | 31.5 | 3.55 (2.90-4.35) | 3.41 (2.73-4.27) | 20.5 | 2.57 (2.07-3.19) | 2.47 (1.95-3.13) | 11.1 | 5.20 (3.45-7.82) | 4.83 (3.29-7.09) |
| Semicontact sports | 19.9 | 1.92 (1.58-2.34) | 2.00 (1.63-2.44) | 14.8 | 1.73 (1.40-2.14) | 1.77 (1.43-2.19) | 5.2 | 2.33 (1.56-3.50) | 2.45 (1.64-3.66) |
| Noncontact sports | 17.9 | 1.70 (1.33-2.16) | 1.65 (1.30-2.09) | 13.7 | 1.58 (1.25-2.02) | 1.55 (1.22-1.96) | 4.2 | 1.83 (1.11-3.03) | 1.77 (1.06-2.96) |
| Other sportsd | 13.3 | 1.19 (0.85-1.65) | 1.11 (0.80-1.54) | 10.0 | 1.11 (0.78-1.58) | 1.06 (0.75-1.51) | 3.3 | 1.47 (0.76-2.28) | 1.33 (0.69-2.58) |
Abbreviations: AOR, adjusted odds ratio; OR, odds ratio.
All models control for sex, race/ethnicity, grade level of respondent, and past-year participation in competitive sports.
Other race includes Asian, American Indian, those who selected multiple races/ethnicities, and those with missing racial information.
For sport participation definitions, see Table 1 footnotes.
Other sports indicates nonspecified other types of sport.
Results
Among the 13 088 adolescents who participated in the 2016 MTF survey, 50.2% were female, 46.8% were white, and the modal age was 16 years (range, 12-18). An estimated 19.5% of adolescents (95% CI, 18.5%-20.6%) reported at least 1 diagnosed concussion in their lifetime (Table 1); 14.0% (95% CI, 13.1%-14.8%) reported 1 diagnosed concussion; and 5.5% (95% CI, 4.9%-6.1%) reported being diagnosed with more than 1 concussion.
Several factors were associated with higher lifetime prevalence of reporting a diagnosed concussion (Table 2): being male, white, in a higher grade, and participating in competitive sports. In particular, participation in contact sports was associated with a higher adjusted odds of lifetime prevalence of being diagnosed with more than 1 concussion (11.1%; AOR, 4.83 [95% CI, 3.29-4.83]).
Discussion
In 2016, 19.5% of US adolescents in grades 8, 10, and 12 reported at least 1 diagnosed concussion during their lifetime, and 5.5% reported being diagnosed with more than 1 concussion. Involvement in competitive sports was associated with a greater odds of lifetime diagnosis of concussions. These findings are consistent with those from emergency department and regional studies that show that participation in sports is one of the leading causes of concussions among adolescents, and that youth involved in contact sports are at an increased risk for sustaining concussions.
The prevalence of concussions among adolescents reported here is much higher than reported from emergency departments (ie, between 4% and 5%), but consistent with regional surveys (ie, 20%). Greater effort to track concussions using large-scale epidemiological data are needed to identify high-risk subpopulations and monitor prevention efforts.
Study limitations include the self-report measure of concussion; responses may be biased due to respondents misunderstanding the question or providing inaccurate information. The cross-sectional design precluded causal or temporal interpretation of the associations found. Temporal ordering of sport participation and diagnosed concussion could not be determined due to the wording of the questions in the survey.
Section Editor: Jody W. Zylke, MD, Deputy Editor.
References
- 1.Ilie G, Boak A, Adlaf EM, Asbridge M, Cusimano MD. Prevalence and correlates of traumatic brain injuries among adolescents. JAMA. 2013;309(24):2550-2552. [DOI] [PubMed] [Google Scholar]
- 2.Ilie G, Mann RE, Hamilton H, et al. . Substance use and related harms among adolescents with traumatic brain injury. J Head Trauma Rehabil. 2015;30(5):293-301. [DOI] [PubMed] [Google Scholar]
- 3.Miech RA, Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. Monitoring the future: mational survey results on drug use: secondary school students. http://www.monitoringthefuture.org/pubs/monographs/mtf-vol1_2015.pdf. Accessed August 1, 2017.
- 4.Halstead ME, Walter KD; Council on Sports Medicine and Fitness . American Academy of Pediatrics. Clinical report—sport-related concussion in children and adolescents. Pediatrics. 2010;126(3):597-615. [DOI] [PubMed] [Google Scholar]
- 5.Comstock DR, Collins CL, Corlette JD, Fletcher EN. RIO: national high school sports-related injury surveillance study: 2011–2012. http://www.nationwidechildrens.org/cirp-rio-study-reports. Accessed February 15, 2013.
- 6.Centers for Disease Control and Prevention TBI data and statistics: emergency department visits, hospitalizations and deaths, 2001–2010. https://www.cdc.gov/traumaticbraininjury/data/index.html. Accessed December 15, 2016.