The relationship of lumbar multifidus muscle morphology to previous, current, and future low back pain: a 9-year population-based prospective cohort study

Jeffrey J Hebert, Per Kjaer, Julie M Fritz, Bruce F Walker

Publikation: Bidrag til tidsskriftTidsskriftsartikelForskningpeer review


STUDY DESIGN: Population based prospective cohort study.

OBJECTIVE: We explored the cross-sectional relationships between lumbar multifidus (LM) intramuscular adipose tissue (IMAT) infiltration and low back pain (LBP) at 3 successive time points and investigated the role of IMAT in predicting the occurrence of LBP after 5 and 9 years.

SUMMARY OF BACKGROUND DATA: Although LBP is a major source of disease burden, the biological determinants of LBP are poorly understood.

METHODS: Participants were 40-year-old adults randomly sampled from a Danish population and followed up at 45 and 49 years of age. At each time point, participants underwent magnetic resonance imaging and reported ever having had LBP, LBP in the previous year, nontrivial LBP in the previous year, or a history of pain radiating into the legs. Pixel intensity and frequencies from T1-weighted magnetic resonance images identified the greatest proportion of LM IMAT at the L4 and L5 spinal levels. IMAT infiltration was categorized as normal/mild, moderate, or severe based on tertile divisions. Associations were explored with crude and adjusted odds ratios (aORs) from logistic regression models. Model covariates included sex, body mass index, and occupational and leisure time physical activity.

RESULTS: A total of 401 participants were enrolled, with 331 (83%) and 286 (71%) participants followed up at 5 and 9 years, respectively. The cross-sectional analyses demonstrated that at the age of 40 years, participants with severe IMAT infiltration demonstrated increased odds of ever experiencing LBP (aOR [95% confidence interval, 95% CI] = 3.16 [1.45-6.89]), nontrivial LBP (aOR [95% CI] = 2.82 [1.36-5.81]), LBP in the past year (aOR [95% CI] = 1.95 [1.07-3.53]), and leg pain (aOR [95% CI] = 2.08 [1.19-3.62]). There were no consistent cross-sectional associations between LBP/leg pain and LM IMAT at 45 or 49 years of age and LM IMAT did not predict future LBP or leg pain.

CONCLUSION: The relationship between LM IMAT and LBP/leg pain is inconsistent and may be modified by age.


TidsskriftSpine (Philadelphia)
Udgave nummer17
Sider (fra-til)1417-1425
Antal sider9
StatusUdgivet - 1 aug. 2014
Udgivet eksterntJa


  • adipose tissue
  • cohort study
  • intramuscular fat
  • low back pain
  • magnetic resonance imaging
  • prediction
  • reliability
  • skeletal muscle