The study protocol was approved by the institutional review board at the National Center for Geriatrics and Gerontology (Approval number 1229), and carries out in accordance with relevant guidelines and regulations. Written informed consent was obtained from all patients.
Patients Enrollment and Eligibility
This observational study was carried out from January 2018 to April 2020 in our institute from a prospectively collected database in the J-BINC. This cohort was a patient-based study that openly recruited individuals aged ≥ 65 years with non-specific chronic pain lasting for more than 6 months, including LBP, neck pain and knee pain. Non-specific CLBP (NCLBP) in this study was defined as follows: (1) LBP with visual analogue scale (VAS) score for LBP ≥ 3; (2) persistent pain localized below the costal margin and above the inferior gluteal folds for more than 6 months; (3) the absence of specific spinal pathology such as infection, tumors, and vertebral fractures on both plain radiographs and lumbar magnetic resonance imaging (MRI); (4) the absence of dominant leg pain caused by radicular and cauda equina disorders; (5) the absence of prominent instability such as spondylolysis, isthmic spondylolisthesis, and degenerative spondylolisthesis more than grade II; (6) no previous lumbar and/or thoracolumbar spine surgery. Degenerative lumbar structures such as the vertebral disc, facet joint, and sacroiliac joint were omitted from the inclusion criteria because available diagnostic procedures for these conditions are inaccurate 15.
Age/sex-matched Control
The retrospective collection was conducted with data from a prospectively maintained database of Sarcopenia Study for Elderly Patient for patients who underwent whole-body dual-energy X-ray absorptiometry (DXA) and evaluated skeletal muscle mass. Registration in this database requires that whole spine radiograph, lumbar MRI, and blood data be performed within 1 year of DXA. Of 2,390 patients (65-100 year, averaged 78.7 years, male 1014 patients, female 1376 patients), 1195 patients excluding lumbar degenerative disease, 683 without complaint of LBP were recruited as control participants.
Laboratory Measurements
Upon enrollment, we collected fasting venous blood samples from patients in the NCLBP and control groups. We recorded complete blood count parameters such as hemoglobin, mean corpuscular volume, white blood cell (WBC) count, lymphocyte count, and red blood cell distribution width (RDW). The RDW is an automated measure of the heterogeneity of red blood cell sizes due to inflammation and senescence of erythropoietic cells in the bone marrow 16.
Radiographic Evaluation
All patients underwent conventional radiography in the standing position. For lateral films, the patients stood with their knees locked, with feet shoulder-width apart, and looking straight ahead. Measured parameters of interest included coronal Cobb angle between the superior edge of L1 and S1, lumbar lordosis (LL), thoracic kyphosis (TK), S1 slope (SS), sagittal vertical axis (SVA), pelvic tilt (PT), pelvic incidence (PI), the presence of spondylolisthesis (anterior slip > 3 mm), and the lumbar range of motion (ROM) defined as the difference in lumbar lordosis angle between flexion and extension. Spinopelvic mismatch was determined when PI-LL is more than 10 ° 17.
Body Composition Analysis
Body composition was assessed using DXA (Lunar iDXA, GE-Healthcare, Tokyo, Japan). Osteoporosis was evaluated using the young adult mean on the lumbar spine (L2-4). Sarcopenia was evaluated using the appendicular lean mass derived from the sum of lean mass in the upper and lower extremities with bone mineral content removed, and skeletal muscle mass index (SMI) was calculated using height squared (kg/m2) 18.
MRI Evaluation
Axial T2-weighted slices at L1/2 and L4/5 were obtained to measure the cross-sectional area (CSA) of the lumbar multifidus and erector spinae muscles for the levels of L1/2 and L4/5. Paraspinal muscle CSAs for both the right and left side were added together for each participant. The CSAs were measured using an area calculation software (SYNAPSER, FUJIFILM MEDICAL, Tokyo, Japan). Vertebral endplate and intervertebral disc degeneration were evaluated based on Modic changes 19 and Pfirrmann classification 20). End plate and disc degeneration were defined as Modic type I, II, and III (except for type 0), and Pfirrmann grade IV and V, respectively.
Statistical Analysis
We determined that a minimum sample of 394 (197 per group) would be required for a power of 90% to detect a clinically importance between-group difference of 0.35 points in the SMI value. Assumptions for the SMI included a two-sided alpha level of 0.05 and a mean standard deviation of 1.07 points 11.
Proportions and means with standard deviations (SD) for normally distributed data and median with minimum and maximum values for not normally distributed data were calculated for covariates and demographic information, moreover, categorical variables were expressed as frequencies or percentages. The chi-square or Fisher exact test was used to assess differences in categorical variables, and means were compared using independent t-test and Mann-Whitney U test for normally and non-normally distributed data, respectively. Normality was checked using the Kolmogorov-Smirnov test. To minimize the effects of potential confounding influences of measured covariates in the 2 study groups (NCLBP vs. control), a propensity score-matched analysis for age and sex was applied. Finally, patients were matched 1:1 without replacement using a nearest-neighbor approach with caliper restrictions set at 0.2. A propensity score was calculated for each patient using the results of this model, regardless of the statistical significance of the independent variables in the model. The correlation between skeletal muscle mass and spinal sagittal alignment parameters were analyzed using simple linear regression analysis (Pearson correlation coefficient). Statistical analyses were performed using the EZR software (Saitama Medical Center, Jichi Medical University, Saitama, Japan). A P value less than 0.05 was considered statistically significant.