Introduction: Pain-influenced and adapted aberrant lumbar motions contain valuable information for clinicians treating patients with chronic low back pain (cLBP). However, objective clinical measurement techniques, like visually recalling multiple simultaneous mechanisms and/or clinical tools (i.e., goniometers, inclinometers, or measurement tape)¹, permit bias and errors to creep in between clinicians and visits. Additionally, valuable multiplanar motions and speed cannot be captured with these techniques. Wearable sensors with inertial measurement units (IMUs) applied to the body have objective and precise biomechanical assessment capabilities of lumbar motions needed for aberrant motion discrimination.^2,3 This current study used portable wireless wearable IMUs to capture test-retest reliability in lumbar ranges of motion (ROM) from participants without cLBP and participants with cLBP on two test days one week apart.

Methods: Participants 18-70 years old were recruited. Asymptomatic inclusion criteria: 1)no LBP history, or 2)most recent LBP episode was two years ago without any symptom return. Chronic LBP inclusion criterion: 1)persistent LBP >3 months. Overall exclusion criteria: 1)cancer history, 2)spinal cord compression, 3)discitis, 4)exercise intolerance, 5)lumbar motion activity restrictions, 6)positional vertigo, or 7)inability to travel to testing facility. Test visits were one week apart. At each visit, participants completed the PROMIS-29+2 patient-reported outcome and were fit with four wearable sensors (Lifeware Labs, Pittsburgh, PA) using double-sided skin adhesive (T1-T2, T12-L1, L5-S1, and on the right lateral thigh approximately 10 cm below the greater trochanter). Participants performed standing flexion, extension, lateral bending, and axial rotation. Triaxial accelerometer, gyroscope, and magnetometer readings were recorded. Post-processing used MATLAB, (The Mathworks, Inc., Natick, MA) and statistical analyses used Minitab v19 (State College, PA), and SPSS v26 (IBM Inc., Armonk, NY).

Results: Twenty-two participants completed all testing. Kolmogorov-Smirnov testing confirmed continuous data were normally distributed. Asymptomatic participants (F6M4, age_ave=43.9yrs(±17.8), 22-67yrs) demonstrated high repeatability on all lumbar ROM and velocity testing (ICC(3,1)=0.73-0.94) and no significant differences between visits (OWANOVA p=.471-.966) (Figure 1-A, B). Participants with cLBP (F10M2, age_ave=41.8yrs(±16.9), 21-69yrs) demonstrated no significant differences between visits (OWANOVA p=.430-.994). PROMIS 29+2 pain intensity ratings (from 0=”no pain” to 10="worst pain imaginable”) from Visit 1 (Pain_ave1=3.6 (±1.4)) to Visit 2(Pain_ave2=3.3(1.2)) were not significant (Mann-Whitney p=.686). Participants with cLBP showed moderate correlation (ICC(3,1) 0.41-0.67) for axial rotation, lateral bending (Figure 1-C), and extension ROM, and axial rotation velocity with high repeatability with lateral bending and flexion velocities (ICC(3,1)=0.84 and 0.71, respectively).

Discussion: We captured highly repeatable IMU data for lumbar ROM from asymptomatic participants demonstrating the IMU sensors are reliable objective tools for measuring lumbar ROM. Additionally, we captured moderate to high ROM repeatability from participants with cLBP. While between-visit pain intensity was insignificant, other outcomes (i.e., pain interference or poor neuromuscular control) may be influencing between-visit variability for those with cLBP. Variables such as motion and speed patterns unique to individuals with cLBP may introduce variability also. Wireless IMUs uniquely capture nuanced motion details unique to individuals. Future goals will capture kinematic data on 1,000 participants with cLBP to characterize lumbar motion patterns coupled with behavioral and biologic data to encompass unique profiles for personalized cLBP treatments.