Fig 4. Biomedical and environmental variables from multivariable regression analyses predicting new injuries.

The thickness of the box represents the number of multivariable models including the following predictors: training characteristics (n = 1, type of training, training frequency, type of terrain); level of competition (n = 2, running pace, running experience); demographics (n = 3, age, gender, BMI); time of the event (n = 2, season); type of the event (n = 1, sport events, carnival/fetes, music concerts).