Quadriceps strength prediction equations in individuals with ligamentous injuries, meniscal injuries and/or osteoarthritis of the knee joint
| aut.embargo | No | en |
| dc.contributor.advisor | McNair, Peter | |
| dc.contributor.advisor | Reid, Duncan | |
| dc.contributor.author | Colvin, Matthew | |
| dc.date.accessioned | 2008-07-15T03:00:31Z | |
| dc.date.available | 2008-07-15T03:00:31Z | |
| dc.date.copyright | 2007 | |
| dc.date.issued | 2007 | |
| dc.description | The objective of this study was to investigate the accuracy of eleven prediction equations and one prediction table when estimating isoinertial knee extension and leg press one repetition maximum (1-RM) performance in subjects with knee injuries and knee osteoarthritis. Study Design: A descriptive quantitative research study was undertaken utilizing a cross-sectional design. Background: Traumatic injuries and osteoarthritis are common musculoskeletal pathologies that can disrupt normal function of the knee joint. A frequent sequela of these pathologies is quadriceps femoris muscle weakness. Such weakness can contribute to disability and diminished levels of functional and recreational activity. Therefore, safe and accurate methods of measuring maximal strength are required to identify and quantify quadriceps strength deficits. One option proposed in the literature is the use of 1-RM prediction equations which estimate 1-RM performance from the number of repetitions completed with sub-maximal loads. These equations have been investigated previously using healthy populations and subjects with calf muscle injuries. However, to date, no known study has investigated their accuracy in individuals with joint pathologies. Method: Machine-weight seated knee extension and seated leg press exercises were investigated in this study. Twenty subjects with knee injuries and 12 subjects with knee OA completed the testing procedures for the knee extension exercise. Nineteen subjects with knee injuries and 18 subjects with knee OA completed the testing procedures for the leg press exercise. All subjects attended the testing venue on three occasions. At the first visit a familiarization session was carried out. At the second and third visits each subject was randomly assigned to perform either actual or predicted 1-RM testing for both of the exercises. Twelve different prediction methods were used to estimate 1-RM performance from the results. The estimates of 1-RM strength were then compared to actual 1-RM performance to assess the level of conformity between these measures. Statistical procedures including Bland and Altman analyses, intraclass correlation coefficients, typical error and total error of measurement were used in the analyses of the results. In addition, paired t-tests were performed to determine whether actual 1-RM values were significantly different across the control and affected limbs and whether there were any significant differences in predictive accuracy for each equation across the control and affected limbs. Finally, the number of subjects with predicted 1-RM values within 5% or less of their actual 1-RM values was determined for each equation. Results: When the knee injury group performed the knee extension exercise, the Brown, Brzycki, Epley, Lander, Mayhew et al., Poliquin and Wathen prediction methods demonstrated the greatest levels of predictive accuracy. When two atypical subjects were identified and excluded from the analyses, the accuracy of these equations improved further. Following the removal of these two subjects, no significant differences in predictive accuracy were found for any of the equations across the affected and control limbs (p > 0.05). Typical errors and total errors were low for the more accurate prediction methods ranging from 2.4-2.8% and from 2.4-3.5%, respectively. Overall, the Poliquin table appeared to be the most accurate prediction method for this sample (affected limbs: bias 0.3 kg, 95% limits of agreement (LOA) -5.8 to 6.4 kg, typical error as a coefficient of variation (COV) 2.4%, total error of measurement (total error) 2.4%; control limbs: bias -1.3 kg, 95% LOA -9.0 to 6.3 kg, typical error as a COV 2.7%, total error 2.8%). When the knee OA group performed the knee extension exercise, the Brown, Brzycki, Epley, Lander, Mayhew et al., Poliquin and Wathen prediction methods demonstrated the greatest levels of predictive accuracy. No significant differences in predictive accuracy were found for any of the equations across the affected and control limbs (p > 0.05). When an atypical subject was identified and excluded from the analyses, the accuracy of the equations improved further. Typical errors as COVs and total errors for the more accurate prediction methods ranged from 2.5-2.7% and from 2.4-2.9%, respectively. Overall, the Poliquin table appeared to be the most accurate prediction method for this sample (affected limbs: bias 0.9 kg, 95% LOA -4.5 to 6.3 kg, typical error as a COV 2.5%, total error 2.5%; control limbs: bias -0.1 kg, 95% LOA -6.0 to 5.9 kg, typical error as a COV 2.5%, total error 2.4%). When the knee injury group performed the leg press, the Adams, Berger, Lombardi and O’Connor equations demonstrated the greatest levels of predictive accuracy. No significant differences in predictive accuracy were found for any of the equations across the affected and control limbs (p > 0.05). Typical errors as COVs and total errors for the more accurate equations ranged from 2.8-3.2% and from 2.9-3.3%, respectively. Overall, the Berger (affected limbs: bias -0.4 kg, 95% LOA -7.2 to 6.3 kg, typical error as a COV 3.2%, total error 3.2%; control limbs: bias 0.1 kg, 95% LOA -6.6 to 6.7 kg, typical error as a COV 3.1%, total error 3.0%) and O’Connor equations (affected limbs: bias -0.6 kg, 95% LOA-6.8 to 5.7 kg, typical error as a COV 2.9%, total error 3.0%; control limbs: bias -0.2 kg, 95% LOA -6.9 to 6.4 kg, typical error as a COV 2.9%, total error 2.9%) appeared to be the most accurate prediction methods for this sample. When the knee OA group performed the leg press, the Adams, Berger, KLW, Lombardi and O’Connor equations demonstrated the greatest levels of predictive accuracy. No significant differences in predictive accuracy were found for any of the equations across the affected and control limbs (p > 0.05). The typical errors as COVs and the total error values for the more accurate prediction methods were the highest observed in this study, ranging from 5.8-6.0% and from 5.7-6.2%, respectively. Overall, the Adams, Berger, KLW and O’Connor equations appeared to be the most accurate prediction methods for this sample. However, it is possible that the predicted leg press 1-RM values produced by the knee OA group might not have matched actual 1-RM values closely enough to be clinically acceptable for some purposes. Conclusion: The findings of the current study suggested that the Poliquin table produced the most accurate estimates of knee extension 1-RM performance for both the knee injury and knee OA groups. In contrast, the Berger and O’Connor equations produced the most accurate estimates of leg press 1-RM performance for the knee injury group, while the Adams, Berger, KLW and O’Connor equations produced the most accurate results for the knee OA group. However, the higher error values observed for the knee OA group suggested that predicted leg press 1-RM performance might not be accurate enough for some clinical purposes. Finally, it can be concluded that no single prediction equation was able to accurately estimate both knee extension and leg press 1-RM performance in subjects with knee injuries and knee OA. | en_US |
| dc.identifier.uri | https://hdl.handle.net/10292/379 | |
| dc.language.iso | en | en_US |
| dc.publisher | Auckland University of Technology | |
| dc.rights.accessrights | OpenAccess | |
| dc.subject | Quadriceps | |
| dc.subject | Strength prediction equations | |
| dc.subject | Knee OA | |
| dc.subject | Knee injuries | |
| dc.title | Quadriceps strength prediction equations in individuals with ligamentous injuries, meniscal injuries and/or osteoarthritis of the knee joint | |
| dc.type | Thesis | |
| thesis.degree.grantor | Auckland University of Technology | |
| thesis.degree.level | Masters Theses | |
| thesis.degree.name | Master of Health Science |