19 The shoulder was passively flexed by the same examiner in all cases, and the end of range was defined

as firm resistance to movement. For the internally and externally rotated positions, the examiner held the shoulder in the respective position prior to flexion, and maintained the position to the end of the available range. The measurements were performed three times for each side, and the mean score was used for further analysis. The data were analyzed using the statistical software package SPSS 15.0 (SPSS Inc., Chicago, IL, USA). Differences in shoulder flexion range of movement between externally and internally rotated positions were analyzed with a factorial analysis of variance (ANOVA) using a mixed model with one between subject factor (sport) and two within subject factors: rotation position (internal and external) and side (left or right). The critical α level was Selleckchem ERK inhibitor 0.05. Paired t tests were used to evaluate specific differences, and Bonferroni corrections were applied. The intra-tester reliability of the measurement technique was assessed by repeating the measurements (shoulder flexion in internal and external rotation) on one shoulder in 10 subjects from the control group, and the intraclass correlation (3,1) and standard error of measurement (SEM) were then calculated,

and showed excellent reliability (r = 0.9, p < 0.01, SEM confidence www.selleckchem.com/products/Neratinib(HKI-272).html interval (95%) = 1.2–1.8°). There were no statistically significant differences in the range of movement for either external or internal rotation (p = 0.78) between sides across all groups. The left and right side data from the each of the three groups were therefore pooled for all further analyses, and a Kolmogorov

Smirmov test demonstrated that the pooled data were normally distributed. The results of the secondly study are shown in Table 1. Analyses of the results using the factorial ANOVA showed that the sport factor had a significant effect on range of movement (p = 0.03), as did the position of rotation (p = 0.001). The interaction of sport and position had no significant effect on range of movement (p = 0.34). Therefore, although the individual sports appeared to have different ranges of movement which differed with shoulder position (internal or external rotation), the relationship between the two positions as defined by the interaction remained fairly constant regardless of the sport. Paired t tests (with Bonferroni corrections) revealed a significant difference between canoeists and swimmers, canoeists and controls, rugby players and canoeists, rugby players and swimmers, controls and swimmers in the ER position (p < 0.017), but not controls and rugby players (p = 0.12). For the IR position, the swimmers differed significantly from the canoeists, rugby players, and controls (p < 0.017), but there were no significant differences between the rugby players, canoeists, and controls (p > 0.07).