Significant differences in surface electromyography (SEMG) have been reported between persons with low back pain (LBP) and normal, healthy controls. This manuscript presents a systematic meta-analytic review of studies examining SEMG differences between these groups. Forty-four articles were identified using MEDLINE and a review of reference lists in articles. For static SEMG, the largest effect size was observed for SEMG while standing, with subjects having LBP demonstrating higher SEMG. The effect size for flexion/relaxation measures was found to be very high (d = -1.71). Studies examining SEMG during isometric exercise or muscle recovery following exercise produced inconsistent findings. Sensitivity and specificity of SEMG for dynamic SEMG measures averaged 88.8% and 81.3%. Most classification schemes were statistically determined and utilized a combination of measures. Only one published study prospectively validated a classification scheme. SEMG measures of flexion-relaxation appear to distinguish LBP subjects from controls with good accuracy, and the sensitivity and specificity of SEMG can be increased by using multiple measures. Further research is needed to determine the combination of measures that are cost-effective, reliable, valid and discriminate with a high degree of accuracy between healthy persons and those with LBP.
Perspective: SEMG is a simple and noninvasive measure of muscle activity. SEMG measures hold promise as an objective marker of LBP.