Background ContextHeterogeneity exists within the low back pain (LBP) population. Some patients recover after every pain episode, whereas others suffer daily from LBP complaints. Until now, studies rarely make a distinction between recurrent low back pain (RLBP) and chronic low back pain (CLBP), although both are characterized by a different clinical picture. Clinical experiences also indicate that heterogeneity exists within the CLBP population. Muscle degeneration, like atrophy, fat infiltration, alterations in muscle fiber type, and altered muscle activity, compromises proper biomechanics and motion of the spinal units in LBP patients. The amount of alterations in muscle structure and muscle function of the paraspinal muscles might be related to the recurrence or chronicity of LBP. PurposeThe aim of this experimental study is to evaluate differences in muscle structure (cross-sectional area and lean muscle fat index) and muscle activity of the multifidus (MF) and erector spinae (ES) during trunk extension, in patients with RLBP, non-continuous CLBP, and continuous CLBP. Study Design and SettingThis cross-sectional study took place in the university hospital of Ghent, Belgium. Muscle structure characteristics and muscle activity were assessed by magnetic resonance imaging (MRI). Patient SampleFifty-five adults with non-specific LBP (24 RLBP in remission, 15 non-continuous CLBP, 16 continuous CLBP) participated in this study. Outcome MeasuresTotal cross-sectional area, muscle cross-sectional area, fat cross-sectional area, lean muscle fat index, T2-rest and T2-shift were assessed. MethodsA T1-weighted Dixon MRI scan was used to evaluate spinal muscle cross-sectional area and fat infiltration in the lumbar MF and ES. Muscle functional MRI was used to evaluate the muscle activity of the lumbar MF and ES during a lumbar extension exercise. Before and after the exercise, a pain assessment was performed. This study was supported by grants from the Special Research Fund of Ghent University (DEF12/AOP/022) without potential conflict of interest-associated biases in the text of the paper. ResultsFat cross-sectional area and lean muscle fat index were significantly higher in MF and ES in continuous CLBP compared with non-continuous CLBP and RLBP (p<.05). No differencesbetween groups were found for total cross-sectional area and muscle cross-sectional area in MF or ES (p>.05). Also, no significant differences between groups for T2-rest were established. T2-shift, however, was significantly lower in MF and ES in RLBP compared with, respectively, non-continuous CLBP and continuous CLBP (p<.05). ConclusionsThese results indicate a higher amount of fat infiltration in the lumbar muscles, in the absence of clear atrophy, in continuous CLBP compared with RLBP. A lower metabolic activity of the lumbar muscles was seen in RLBP replicating a relative lower intensity in contractions performed by the lumbar muscles in RLBP compared with non-continuous and continuous CLBP. In conclusion, RLBP differs from continuous CLBP for both muscle structure and muscle function, whereas non-continuous CLBP seems comparable with RLBP for lumbar muscle structure and with continuous CLBP for lumbar muscle function. These results underline the differences in muscle structure and muscle function between different LBP populations.