Test-retest reliability of the soleus H-reflex in three different positions.

PURPOSE: H-reflex has been clinically useful in the diagnosis of radiculopathies, developmental disorders, and measurement of motoneuron excitability. However, variability of the H-reflex precluded its routine application. The purpose of this study is to evaluate the test-retest and within-subject reliability of the soleus H-reflex tested in three different positions. SUBJECTS: Seven men and eight women healthy volunteered (20-50 y) with no history of significant low back pain or radiculopathy consented to the study. METHODS: The soleus H-reflexes for both lower extremities were elicited and recorded using Cadwell 5200-A EMG unit and surface recording. The tibial nerve was electrically stimulated at the popliteal fossa using 0-5 ms., 0.2 pps pulses at intensity equivalent to H-max. Each subject was tested randomly in three different positions: pronelying, free standing, and standing while lifting 20% of his/her body weight. Signal were amplified (1-5 K) using surface electrodes applied on the soleus muscle at midline and 3 cm below the gastrocnemius musculotendinous junction. The peak-to-peak amplitude and onset latencies of four separate traces were averaged for each trial. Subjects were re-tested within 10 days by the same tester following the same protocol. RESULTS: Test-retest reliability of the H-reflex amplitude ranged from r = .29 in prone position to r = .56 in the loading position. Within day reliability of the H-amplitude was high between the three different positions and ranged from r = .56 to r = .97. The test-retest reliability of the H-latency were extremely high and robust, with the coefficients ranged from r = .92 to r = .94. Also the within day reliability of the H-latency ranged from r = .96 to r = .99. CONCLUSIONS: Results indicated that, when the H-amplitude is the measure of choice, testing the H-reflex in standing and loading positions is more reliable than testing in pronelying. Also testing the subject during various procedures in the same session is more reliable than testing subject in different days/sessions. The H-latency is highly reliable in all three testing positions.