Predicting Clinical Improvement in Chronic Low Back Pain Using Exercise‐Induced IVIM Perfusion: Threshold Evaluation and Disability‐Dependent Effects

ABSTRACT Chronic low back pain (LBP) is the leading cause of disability worldwide, yet responses to exercise‐based rehabilitation vary substantially and remain not well characterized. Impaired activation of the lumbar paraspinal muscles has been proposed as a mechanism limiting rehabilitation treatment response. Intravoxel incoherent motion (IVIM) magnetic resonance imaging assesses muscle diffusion and perfusion after exercise, offering physiological markers of dynamic function. In a previous pilot study, we reported that acute exercise‐induced IVIM responses were reduced in individuals with LBP and that thresholds in diffusion (D) and pseudodiffusion (D*) parameters could predict those who would or would not respond to an exercise‐based rehabilitation program. The present study aimed to prospectively validate these findings in a larger, independent cohort. Thirty‐seven individuals with chronic LBP underwent IVIM before and after an acute standardized lumbar extension exercise, followed by a 12‐week rehabilitation program. Participants were classified as responders or nonresponders based on changes in disability after 12 weeks (Δ Oswestry Disability Index Score ≥ or < 0). In line with the pilot study, responders exhibited higher increase in D following acute exercise than nonresponders ( p = 0.05). Previously identified thresholds for D and D* did not discriminate between responders and nonresponders within this new cohort (Sn < 0.40 and Sp = 0.65). Exploratory analyses revealed that baseline disability moderated IVIM responses, as a marked increase in D was primarily observed in responders with low baseline disability, whereas individuals with higher disability showed attenuated responses. New significant thresholds were explored within the validation cohort while controlling for disability levels. An increase ≥ 0.014 × 10 −3 mm 2 /s for D predicted who would respond to the exercise‐based program, with a Sn of 0.59 and a Sp of 0.90 (area under the curve [AUC] 0.72, 95%CI 0.55 to 0.90, p = 0.02). These findings confirm that acute IVIM responses are associated with rehabilitation outcomes but may be better understood as part of a broader, integrative framework characterizing the physiological mechanisms associated with treatment response. Integrating physiological and clinical factors to better understand variability in treatment response will help guide more personalized approaches to LBP rehabilitation.