TABLE 2—
Past-Year Nonmedical Use of Prescription Opioids Among 8th and 10th Graders in the United States, by Sports Participation: Monitoring the Future, 2011
| Sporta | AOR (99% CI) |
| Baseball | 1.20 (0.93, 1.55) |
| Basketball | 0.81 (0.62, 1.05) |
| Cheerleading | 1.37 (0.92, 2.04) |
| Crew | 0.72 (0.25, 2.09) |
| Cross country | 0.45 (0.20, 1.00) |
| Equestrian | 1.50 (0.66, 3.43) |
| Field hockey | 1.30 (0.60, 2.82) |
| Football | 1.50 (1.12, 1.99) |
| Golf | 1.38 (0.88, 2.16) |
| Gymnastics | 1.34 (0.79, 2.28) |
| Ice hockey | 0.88 (0.44, 1.78) |
| Lacrosse | 1.03 (0.56, 1.89) |
| Soccer | 1.04 (0.77, 1.39) |
| Swimming | 0.91 (0.61, 1.34) |
| Tennis | 0.85 (0.53, 1.37) |
| Track and field | 1.05 (0.77, 1.44) |
| Volleyball | 1.09 (0.79, 1.51) |
| Water polo | 1.28 (0.48, 3.40) |
| Wrestling | 1.49 (1.01, 2.19) |
Note. AOR = adjusted odds ratio; CI = confidence interval. AORs control for year of survey, region and geographic location of respondent, family characteristics, and student characteristics simultaneously. AORs and CIs for control variables are not shown. Missing cases were excluded from the analysis, leaving a final sample size of 13 751 adolescents. Cox and Snell R2 = 0.057.
Each sport is a separate binary variable (i.e., 1 indicates that they participate in a given sport and 0 indicates that they do not participate in a given sport) and are all included in the logistic regression model with the control variables simultaneously. Inclusion of all 19 sports into the model simultaneously controls for number and type of sports that respondents participate in.