Reproducibility of corticomotor threshold: Some observations

Recent articles by Mills et al. 2,3 raise the important question of achieving reproducible threshold (T) measurements in clinical and physiological studies. A procedure for determining threshold values was defined by an International Committee, 4 which advocates increasing stimulus intensity in 5% steps until the level T is reached which induces motor responses in about 50% of 10–20 consecutive stimuli (Method I). Mills and Nithi 2 propose an alternative procedure in which the T region is scanned in 1% steps, by using a figure eight-shaped coil, to define a lower T (LT): 0 responses in 10 stimuli, and an upper T (UT): 10 responses in 10 stimuli (Method II). The two procedures differ not only in the stepping value but also in the points where the T curve is tested. The last aspect could affect the reproducibility of T measurements if the slope of the T curve, idealized as a plot of the probability of obtaining a response versus stimulus intensity, is different at its center (where T is measured) and at its end points (where LT and UT are measured). In order to investigate this point we measured T, LT, and UT for both abductor digiti minimi muscles on 8 healthy right-handed subjects (7 women, 27–49 years old), with their informed consent. Recordings were made with surface electrodes using a belly-tendon montage. Signals were recorded by a Keypoint EMG system (Dantec) with filters set to pass 20 Hz to 10 kHz and a sensitivity set at 100 µV/division. All measurements were carried out using steps of 5% in stimulus intensity. Biphasic stimulation was applied over the vertex with a 13-cm round coil (MagPro, Dantec) with the muscle relaxed. One investigator, who was blind to stimulus intensity, performed all stimulations and decided whether or not a response had been obtained. Mean T values are shown in Table I. Standard deviations are similar for both methods. These spreads in T values can be ascribed to two different sources of variation: differences between subjects and variation in measurements carried out on the same subject. An estimate of the within-subject variability was obtained by analysis of variance , in the form of an estimate of the standard deviation within subjects, (MS within) 1 ⁄2. This figure determines the reproducibility of individual measurements. The test– retest repeatability 2 is based on this figure and is also similar for both methods (see Table 1). The …