The incidence of low-back pain (LBP)is high among elite male high school soft tennis players. A previous report showed that hip internal-rotation (IR) restriction in the non-dominant leg could increase the risk of LBP. Moreover, IR in the dominant shoulder is important for serve and stroke movements, suggesting that IR restriction in the dominant shoulder can contribute to LBP. The simultaneous incidence of non-dominant-hip IR restriction may help in developing a good LBP-related factor model; therefore, this study aimed to investigate the association between dominant-shoulder IR restriction and LBP in elite male high school soft tennis players. This is a single-center cross-sectional study. This study included 176 male high school soft tennis players from 14 elite teams. The main measures were IR and external-rotation range of motion (ROM) in the dominant and non-dominant shoulders and hip, assessed using a goniometer with a bubble attached. Multiple logistic regression analyses were performed with the presence of LBP as a categorical variable (LBP, 1; non-LBP (NLBP), 0). Multiple logistic regression models were created as follows: Model 1 included dominant-shoulder IR restriction (loss of glenohumeral IR (GIRLoss)), Model 2 included non-dominant-hip IR ROM, and Model 3 included both GIRLoss and non-dominant-hip IR ROM as the main explanatory variables to test the association between dominant-shoulder IR restriction and LBP, compare its suitability as an associated factor with non-dominant-hip IR restriction, and determine if either or both should be considered in an LBP association model based on model fit. The significance of each multiple logistic regression model was examined using the model χ2 test, and a model with P<0.05 was judged as a significant model. Model fit was examined using the Hosmer-Lemeshow test, and P≥0.05 was judged as a good model fit. The Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were used to compare the goodness-of-fit among multiple logistic regression models. Results: The LBP and non-LBP groups comprised 59 and 117 players, respectively. GIRLoss in the dominant shoulder was a significant independent factor (odds ratio: 1.06, 95% CI: 1.02-1.09; P<0.01) in Model 3, which included hip IR restriction in the non-dominant leg (odds ratio: 0.90, 95% CI: 0.86-0.94; P<0.01).In all models, the explanatory variables fitted were significantly associated with LBP, indicating that the models were significant and fitted well: the AICs (and BICs) for Models 1, 2, and 3 were 198.4 (207.6), 178.6 (187.9), and 168.8 (181.1), respectively, indicating that Model 3 was the best fit. In elite male high school soft tennis players, IR restriction in the dominant shoulder was associated with LBP. IR restriction in the dominant shoulder and non-dominant hip as a significant factor for LBP may contribute to developing an optimal multiple logistic regression model for LBP.
Read full abstract