Dingbo Shu, Jianping Wang, Haijuan Wu, Yanghua Chen, Li Zhao, Zhenhua Zhao
After FDR correction, ADC was associated with Modic changes (OR = 1.520), endplate defects (OR = 2.072), and nerve root compression (OR = 1.641), while PDFF was associated with central canal stenosis (OR = 2.152). Age (standardized regression coefficient (β) = 0.534) and BMI (β = 0.200), but not ADC (β = 0.182, q = 0.123), were associated with degenerative burden. No MRI parameter was associated with low back pain after FDR correction (q = 0.312-0.975). Three exploratory clusters differed in age and sex, but not adjusted degenerative burden (nominal p = 0.381).
BACKGROUND: Vertebral marrow alterations are associated with lumbar spine degeneration, but their relationships with marrow fat and diffusion measurements remain unclear.
PURPOSE: To assess associations of vertebral marrow proton density fat fraction (PDFF), apparent diffusion coefficient (ADC), slow diffusion coefficient (SDC), and diffusion-derived vessel density (DDVD) with lumbar structural degeneration and low back pain.
STUDY TYPE: Prospective cross-sectional study.
POPULATION: One hundred and sixty-two community-dwelling adults (male/female: 71/91, age: 55.1 ± 11.1 years, and body mass index (BMI): 24.4 ± 3.3 kg/m2).
FIELD STRENGTH/SEQUENCE: 3.0-T; sagittal multi-b-value readout-segmented echo-planar DWI and transverse six-echo three-dimensional q-Dixon imaging.
ASSESSMENT: PDFF, ADC, SDC, and DDVD were measured adjacent to the L3/4-L5/S1 endplates. Six structural features were independently evaluated by two radiologists with 8 and 7 years of experience and summed across levels as a degenerative burden score (range, 0-18). Chronic low back pain was self-reported pain lasting at least 3 months.
STATISTICAL TESTS: Generalized estimating equations, multivariable linear and logistic regression, principal component analysis, and K-means clustering. p < 0.05 and FDR-adjusted q < 0.05 were considered statistically significant.
RESULTS: After FDR correction, ADC was associated with Modic changes (OR = 1.520), endplate defects (OR = 2.072), and nerve root compression (OR = 1.641), while PDFF was associated with central canal stenosis (OR = 2.152). Age (standardized regression coefficient (β) = 0.534) and BMI (β = 0.200), but not ADC (β = 0.182, q = 0.123), were associated with degenerative burden. No MRI parameter was associated with low back pain after FDR correction (q = 0.312-0.975). Three exploratory clusters differed in age and sex, but not adjusted degenerative burden (nominal p = 0.381).
DATA CONCLUSION: ADC showed the most consistent adjusted associations with lumbar structural degeneration. The clinical significance of the identified vertebral marrow imaging profiles remains uncertain.
EVIDENCE LEVEL: 3.
TECHNICAL EFFICACY: Stage 2.