Person-days of enrollment at UNCSA were calculated based on the length of academic terms at UNCSA.
44-47) For our study, data on potential predictors of injury were obtained from two sources: a pre-matriculation student health questionnaire and anthropometric assessments conducted by the UNCSA SHS at time of matriculation.
Measurements to calculate BMI and BF% were collected by UNCSA SHS staff when each dancer enrolled at UNCSA.
The questionnaire was used by UNCSA SHS for administrative purposes and included measures of mental health history, risk-taking behavior, and injury history.
as a dichotomous variable indicating whether a student was injured during their previous semester at UNCSA and 2.
For the time-varying covariates of age and history of injury at UNCSA, variables were parameterized both as continuous and categorical variables in all steps to determine which was the best fit for the models.
There were 480 dancers at UNCSA who began enrollment during or after Fall 2009 and were followed up through either the end of their studies at UNCSA or Spring 2015.
The maximum number of injuries sustained by a UNCSA dancer was 15 (this dancer was followed for 2.
14: reference = 15 or younger), number of injuries sustained at UNCSA prior to the semester of current injury (1 injury IRR: 1.
Using multivariable modeling, we found that the best predictive model (based on the low QICu score) for females was a self-reported history of depression, age at time of injury, number of injuries sustained at UNCSA prior to the current semester, BMI, dance style, and history of back pain.
We also limited the likelihood that loss-to-follow-up biased our results by only including dancers we could observe from the beginning of their enrollment at UNCSA.