ORIGINAL ARTICLE Veerle Hermans · Arthur J. Spaepen Influence of electrode position on changes in electromyograph parameters of the upper trapezius muscle during submaximal sustained contractions Accepted: 13 August 1996 Abstract The purpose of the present study was to eval- uate whether differences could be found in the changes in mean power frequency (MPF) and root-mean-square (rms) due to electrode positions on the upper trapezius muscle during a sustained submaximal task. A group of 25 healthy subjects performed a continuous forward flexion of the right arm at 20% of their maximal vo- luntary contraction (MVC). Three pairs of bipolar sur- face electrodes were positioned on a straight line be- tween the spine of the seventh cervical vertebra (C7) and the lateral edge of the acromion. The fourth pair (the caudal position) was placed 2 cm below the midpoint of this line. The analysis of variance revealed significant differences between the electrode positions during the short 20% MVC and during MVC (P ≤ 0.01). Further- more, lower rms values were found when the electrodes were placed symmetrically around the midpoint of the reference line, which confirms previous studies. A sta- tistical model has been developed to discriminate ad- ditionally the very small changes in the electromyogram parameters due to electrode position during the sus- tained 20% MVC. It was found that the slope coeffi- cients of rms and MPF were significantly different from 0(P ≤ 0.01), although to a lesser degree for rms and more clearly for MPF of the caudal position (P ≤ 0.05). Furthermore, significant differences were found between several combinations of the uppermost electrodes and the caudal position (P ≤ 0.01). From this study, it is concluded that it is important to evaluate several elec- trode positions on the upper trapezius muscle to be able to represent the behaviour of this muscle accurately during a sustained contraction. Key words Electromyography · Endurance · Trapezius muscle · Electrode position Introduction Muscle activity during occupational tasks is often in- vestigated using surface electromyograph recordings (EMG) of the shoulder complex (Jonsson 1978; West- gaard 1988). Attention has been focused on the upper trapezius muscle since this muscle is a prevalent site of work-related pain and disorders, often referred to as chronic trapezius myalgia (Hagberg 1984; Larsson et al. 1988; Lindman et al. 1990, 1991). The recommendations of Zipp (1982) for positioning the bipolar surface elec- trodes on the trapezius muscle should contribute to- wards good signal repeatability and take into consideration individual body dimensions: the electrodes should be placed symmetrically around the midpoint of the connecting line, or lead line, between cervical ver- tebra seven and the lateral part of the acromion. Although many recent studies still use this position (e.g. Takala et al. 1993; Nieminen et al. 1993; O ¨ berg et al. 1994; Larsson et al. 1995), it has been found that the recording of a signal 2-cm lateral of the midpoint not only gave a higher signal, but also showed a higher repeat consistency or lower variation (Veiersted 1991; Jensen et al. 1993). The larger variation in EMG amplitude at the midpoint has been explained by the location of the bi- polar electrodes symmetrically around the innervation zone and therefore the more lateral position has been considered as the most suitable location for representing the upper trapezius and for measurement of shoulder muscle activity in laboratory or occupational settings. Apart from evaluating occupational load, it is also important to study to what extent the mentioned influ- ence of electrode position on the EMG signal also occurs during sustained, submaximal tasks. It is known that changes in the electromyograph signal are found during these activities which can be quantified using parameters from the time and the frequency domain of the signal. An increase in amplitude and a decrease in the power frequency have been considered as indications of muscle fatigue (De Luca 1984). Mathiassen (1993) has in- Eur J Appl Physiol (1997) 75: 319–325  Springer-Verlag 1997 V.Hermans (&) · A.J. Spaepen Laboratory of Ergonomics and Occupational Biomechanics, Faculty of Physical Education and Physiotherapy, Katholieke Universiteit, Tervuursevest 101, B-3001 Leuven, Belgium