RESUMO
BACKGROUND: Patients with chronic nonspecific low-back pain (CNSLBP) lack the flexion-relaxation phenomenon in full-trunk bending. This can be quantified by surface electromyography (SEMG) measurement of lumbar erector muscle. The study objective is to explore the clinical utility of the SEMG flexion-extension ratio (FER) in distinguishing patients with CNSLBP from painfree persons. METHODS: This was a comparative cross-sectional study. We adopted a balanced study design by recruiting 130 participants each for the CNSLBP and control arms. Each participant underwent dynamic SEMG measurement in full-trunk bending, which consisted of standing, flexion, relaxation, and extension. The FER ratio was the ratio of the maximum SEMG in flexion to the maximum SEMG during extension. Receiver-operating characteristic (ROC) analysis was conducted to identify optimal values of the FER and associated sensitivity, specificity, and diagnostic accuracy. RESULTS: The CNSLBP group and control group were generally comparable in terms of demographics and clinical profile. The CNSLBP group had higher SEMG amplitudes during flexion but lower SEMG during extension. The mean (SD) FER of the CNSLBP group was 0.90 (0.26), which was almost double that of controls 0.47 (0.14). The ROC curve identified an optimal FER cutoff of ≥ 0.692, for which sensitivity and specificity were 76.15% (95%confidence interval [CI], 68.14-82.66) and 98.46% (95%CI, 94.56-99.58). The diagnostic accuracy was 92.1% (95%CI, 88.70-95.54). CONCLUSION: The FER derived by lumbar muscle SEMG is able to distinguish patients with CNSLBP from pain-free people with excellent accuracy. This provides good evidence that a customized FER can be used in various clinical scenarios.