Evaluation of cervical proprioceptive function: optimizing protocols and comparison between tests in normal subjects

STUDY DESIGN: A test-retest design evaluated stability as well as within and between day reliability. OBJECTIVE: The study aimed to define optimum protocols for the cervical joint position error (JPE) and cervicocephalic kinesthesia tests and to investigate association between performances in the tests. SUMMARY OF BACKGROUND DATA: The cervical JPE and cervicocephalic kinesthesia tests are proposed as measures of cervical proprioception. However, there has been little investigation of the number of trials needed to obtain stable and reliable estimates of performance. Both tests have potential limitations in reflecting the underlying construct of cervical proprioception and association between performances in both has not been investigated previously. METHODS: Head repositioning and head-tracking errors were measured using an electromagnetic-tracking system in 16 normal subjects, tested on 3 occasions over 2 days. The effect of different numbers of trial repeats was analyzed descriptively in terms of stability of measures obtained and by using intraclass correlation coefficients to assess reliability. Association between the tests was analyzed with the Pearson correlation coefficient. RESULTS: Stable estimates of performance were obtained when data from 6 or more trials was included. The greatest test-retest reliability was obtained with 5 or more trials in both the cervical JPE (intraclass correlation coefficients = 0.73-0.84) and cervicocephalic kinesthesia (intraclass correlation coefficients = 0.90-0.97) tests. Correlation analyses indicated no significant association between performances in the 2 tests (r = -0.476-0.228, P > 0.05). CONCLUSION: Our finding that at least 6 trials were needed to optimize stability, and reliability of outcome measures has important implications for application of these tests. The lack of correlation between performances in the tests supports the suggestion that they are not comparable measures of cervical proprioception. Further planned studies will include a range of tests challenging different aspects of cervical proprioceptive contribution to sensorimotor control in different subcategories of neck pain patients.