Higher lumbodorsal epimuscular fat associated with recurrent low back pain: an exploratory secondary analysis of cross-sectional MRI data with 1-year follow-up.

PURPOSE: To explore segment-specific associations between lumbodorsal epimuscular fat (EMF FF%) and paraspinal intramuscular fat (IMF FF%) fractions and recurrent low back pain (LBP) within one year, to assess whether FF% may be linked to recurrence risk, and to examine correlations between EMF FF% and paraspinal IMF FF% or muscle tissue fraction (MT%).

METHODS

This secondary analysis included 76 adults who had recently recovered from acute LBP (≥ 3 months) and were symptom-free at baseline MRI. MRI scans were acquired between 2012 and 2013. EMF FF%, IMF FF%, and MT% were quantified from L1–S1 using Gaussian mixture modeling. Recurrent LBP (yes/no) was assessed at one-year follow-up during consultations. Binary logistic regression (adjusted for age, sex, BMI) and relative risk (RR) analyses were used to assess recurrence on complete cases. Multiple linear regression with post hoc partial correlations examined segmental EMF FF%–IMF FF%/MT% associations. Multicollinearity was assessed and addressed by removing highly correlated predictors.

RESULTS

Higher EMF FF% at L1–L2 was associated with increased odds of recurrence (Exp(B) = 1.579, CI95%: 1.01–2.48, p = 0.047), while EMF FF% at L2–L3 was associated with decreased odds (Exp(B) = 0.686, CI95%: 0.49–0.95, p = 0.025). Higher LM IMF FF% at L3–L4 was associated with increased odds (Exp(B) = 1.166, CI95%: 1.03–1.32, p = 0.011). Adjusted RRs were non-significant. EMF FF% and IMF/MT% showed segment-specific correlations, e.g., negative associations between EMF FF% and adjacent ES MT%.

CONCLUSION

Fat distribution in the lumbar region is associated with recurrent LBP within one year in a segment- and location-specific manner. Higher EMF FF% at L1–L2 and IMF FF% in the multifidus at L3–L4 increased odds of recurrence, whereas higher EMF FF% at L2–L3 was associated with lower odds. These findings highlight the importance of considering both fat quantity and location, as well as interactions with paraspinal muscle composition, in evaluating recurrence risk. Future studies should explore these relationships in larger cohorts.