Previous studies of EMG recordings from lumbar para-spinal muscles have shown correlations between some EMG variables and low back pain. However there are discrepancies in the literature concerning the usefulness of some of these variables. It has been suggested that ordinary fatigue influences the reproducibility of these measurements, introducing a confounding factor.

In this study we have investigated changes in EMG variables, following a day of normal activity. Forty six subjects participated in this study. EMG recordings were performed at the beginning of their shift (time 1) and at 6 h 20 ±5 min afterwards (time 2) under isometric condition at 60% and 40% of their lean body mass (LBM). Variables studied were initial medial frequency (IMF), median frequency slope (MFS) and half width (HW).

At 60% LBM, IMF measurements at time 1 and time 2 were highly correlated (r2= 0.84, p> 0001) and this was the case for HW measurements (r2=0.84, p> 0001) and MF slope (r2=0.52, p=0> 001). Conducting paired sample t-test also showed no significant change in the IMF from time 1 (M=48.6, SD=8.9) to time 2 (M=49.2, SD=7.3), t(45)=−0.9, p=0.38, or in HW from time 1 (M=47.2, SD=15.5) to time 2 (M=45.9, SD=13.9), t(45)=1.7, p=0.29, or MF slope from time 1 (M=−0.2, SD=0.17) to time 2 (M=−0.24, SD=0.16), t(45)=1.67, p=0.10). The relations observed at 40% LBM almost mirrored those reported at 60 % LBM but with even less significant difference from time1 to time2.

We conclude that IMF, HW and MFS are highly reproducible EMG variables that are not affected by ordinary fatigue and may therefore be valuable in examining differences between subjects or over longer time periods. However they are not useful in assessing changes due to daily exertion.

Purpose: Low frequency peaks (LFP) commonly observed in EMG spectra of paraspinal muscles. These peaks have frequency of 11–15Hz (Median=13, SD=4) and commonly observed in 40% of EMG recordings. We examined the correlation of these peaks with variation in within individuals change in load, between individuals load, gender, age, history of back pain and HW.

Methods and results: Nineteen healthy volunteers were recruited for study of within subject variation in load and data from 106 subjects was examined for the remaining analysis reported here. EMG acquisition was performed using the method described by Oliver et al (Oliver et al., 1996) at a range of load varying from a kilogram to 100% MVC in increments of 10% MVC. The tests were performed in a random order. This method was used for acquisition of data from the second group (n=106) with an exception that only one load fixed at 2/3MVC was used. All parameters including Low frequency peaks and Half Width (HW) was calculated by an automated software developed for this purpose

No correlation between within individual change in load and presence or absence of LFP was found. Only one subject showed a significant correlation between individual change in load with the magnitude of LFPs (r=−0.75, p=0.012). However, when the cumulative data from all 19 subjects was analysed, small correlation between the change in load and magnitude of LFPs (r=−0.17, n=187, p=0.022) was observed, with greater load associated with smaller LFP. No other correlation between gender, individuals load, age and fitness was found.

Conclusion: The appearance of the LFP has been previously described. As yet the underlying cause is not understood, but may be related to a load/feedback loop. The results of this study suggest that load is a factor that may affect LFP and this should be taken into consideration when methods such as HW are used.

Introduction: EMG recording from lumber spine muscles can be a reliable discriminator and predictor of low back pain (LBP). Multi variant analysis shows age influences these variables.

Aim: to determine in a longitudinal study if age is a significant factor.

Method: EMG recording from 9 subjects was carried out at time1 and 12 years afterwards (time 2);

at the same load and

Results: At the same load there was no significant change in the Mean Half Width (HW) from time1 (M=45.9, SD=19.1) to time 2 (M=51.4, SD=18.7), t(8)= −0.98, p=0.36. No difference was observed when the load set at time 1 was used at time 2 (M=51.4, SD=18.7) and compared with a load set from the MVC obtained at time 2 (M=45.9, SD=12.0), t(8)=1.75, p=0.118.. There was no statistically significant difference between Initial Median Frequency (IMF) at time1 (M=50.6, SD=12.0) and time 2, either using the same load (M=51.7, SD=8.6), t(8)=− 0.273, p=0.79) or the load based on current MVC.

Discussion: In this 12 year longitudinal study, age did not appear to affect the HW or IMF measurement. Both of these variables might be used in long term studies.