Subcutaneous fat thickness is not correlated to the magnitude of measurable trunk muscle activity in surface EMG of normal weight healthy adults

Introduction and Objectives: Surface Electromyography (sEMG) is a non-invasive tool from which muscle activation parameters can be investigated during a variety of static and dynamic conditions. However, subcutaneous fat overlying a muscle of interest can reduce the amplitude of the signal obtained, as fat can act as a low pass filter [1,2]. The effect of subcutaneous tissue and skinfold thickness on measurements of muscle activity has been evaluated in previous studies using ultrasound and calipers [2,3]. However to our knowledge, magnetic resonance imaging (MRI) assessment, which is a more precise method of determining subcutaneous fat thickness, has not been compared to sEMG values of trunk muscle activity. Therefore, the aim of the study was to compare the sEMG output of two trunk muscles with skin/subcutaneous fat thickness determined from MRI. Methods: From data obtained previously, EMG and MRI values of two separate measurement sessions from 12 healthy participants aged 18-25 years (female n=6) with a BMI of <25 were correlated. Muscle activity had been obtained through sEMG electrodes (Delsys Trigno, 1200Hz) attached over the left and right erector spinae (ES) 2cm laterally from the thoracolumbar junction, and rectus abdominis (RA) 2cm above the umbilicus and 1cm lateral to midline. Data had been collected during 3 trials of 10s at natural stance. For the present study, raw sEMG data were processed using two methods: (1) resampling, rectification, zero-mean offset, and a 4th order low pass Butterworth filter applied, followed by the mean activity being calculated from the middle 6s of each 10s trial, and (2) EMG integration by cumulative summation of the rectified values. For the MRI aspect of the study, transverse images were obtained from the distal thoracic spine to the pelvis using a T1-weighted sequence (1.5T Philips Achieva; slice thickness 10mm, gap 1mm, repetition time 9.3ms, echo time 4.6ms, flip angle 15 degrees, field of view (FOV) 345mm, rectangular FOV 78%). Total skin and subcutaneous fat thickness (termed SC) of the abdomen was measured at L2-L3 disc level, equating to the RA sEMG site, and SC of the back was measured at T12-L1 disc level equating to the ES sEMG site (Sante DICOM viewer freeware, v4.0.5). For both locations, this was performed for the left and right side separately. Normal distribution of the data was confirmed using a Shapiro-Wilk test and Pearson’s correlations were carried out between the total thickness of skin and subcutaneous fat and the respective sEMG value for each muscle, separately for each EMG processing method. Results: Mean BMI of the participants was 21.7 ± 2.2. The means and standard deviations of the MRI and EMG data are shown in Table 1. Negative correlations were found between the SC values and EMG output in all muscles and from both EMG processing methods, although none of these findings were significant (all p>0.05). Conclusion: The findings from the present study show that, although an inverse relationship i.e. negative correlation between SC and sEMG values was expected, this was not significant in either of the two muscles. Therefore, in healthy people who are not classed as being overweight, the effects of skin/subcutaneous fat thickness on resultant measurable muscle activity with sEMG can be considered negligible in static, minimal effort conditions, based on the two different EMG processing methods described here. Identifying whether this also applies during dynamic and isometric contractions should be further investigated, however during dynamic activities the effect of the increased skin displacement with increased subcutaneous fat thickness should be considered.