Abstract
-
Objective
Low back pain (LBP) is common, yet many individuals maintain normal activities of daily living despite chronic symptoms and structural changes evident on imaging. We hypothesized that functional resilience, defined as preserved functional capacity despite pain and age‑typical degenerative changes, represents a meaningful clinical phenotype, and that function‑centered outcome measures would better discriminate disability status than structural imaging features.
-
Methods
This study analyzed 347 participants reporting LBP from the Wakayama Spine Study (N=866). Maintained function was defined a priori as Oswestry Disability Index (ODI) ≤20%. We compared those with maintained function (n=220, 63.4%) to those with impairment (n=127) across demographics, lifestyle, metabolic components, physical performance (grip strength, gait speed), and lumbar magnetic resonance imaging (MRI) findings. Multivariable logistic regression among participants with LBP, including age, sex, obesity, metabolic factors, pain intensity, physical performance, and MRI phenotypes, was used to identify independent predictors of functional resilience.
-
Results
Functional resilience was common: 63.0% of LBP participants had ODI ≤20%. Resilient individuals were younger (65.0±11.9 years vs. 74.6±10.9 years, p<0.001) with superior physical performance. In multivariable models, male sex predicted maintained function (odds ratio [OR], 1.76; 95% confidence interval [CI], 1.03–3.00; p<0.05), while obesity (body mass index ≥25 kg/m2) was associated with reduced odds of resilience (OR, 0.50; 95% CI, 0.30–0.84; p<0.01). Standard MRI features, including disc degeneration, Modic changes, and Schmorl nodes, were not independently associated with functional status after adjustment, despite disc degeneration being highly prevalent even among resilient participants (95.4%).
-
Conclusion
These data confirm that functional resilience is common in LBP and is not negated by the presence of structural MRI abnormalities. Among LBP patients, male sex and absence of obesity are independent predictors of maintained function, whereas standard MRI features do not independently predict functional status after age adjustment. Function-centered metrics (ODI, gait speed, grip strength) better discriminate functional status than structural imaging findings.
-
Keywords: Low back pain, Functional resilience, Activities of daily living, Obesity, Disc degeneration, Normal activities of daily living
INTRODUCTION
Low back pain (LBP) represents one of the most prevalent musculoskeletal conditions globally [
1-
12], affecting approximately 540 million people worldwide at any given time [
3]. The Global Burden of Disease Study 2019 identified LBP as the leading cause of years lived with disability in most countries [
4], with estimated lifetime prevalence rates ranging from 60% to 85% [
5,
6]. This high prevalence translates into substantial socioeconomic impact, with annual direct medical costs estimated at $90 billion in the United States alone [
7]. Yet the field still struggles with a fundamental methodological problem: what constitutes an appropriate “normal” or control state.
The conventional clinical paradigm views LBP primarily as a source of functional disability [
8-
10]. However, this perspective has been increasingly challenged by observations that pain intensity correlates poorly with functional impairment [
11,
12]. Many studies implicitly equate normality with the absence of pain and the absence of structural changes on imaging, an idealized baseline that is rarely observed in real-world populations, especially with advancing age. However, some individuals maintain normal activities of daily living (ADL) despite persistent pain, suggesting that factors beyond pain intensity influence functional outcomes [
11-
14]. This phenomenon has led to the emergence of “functional resilience” as a concept in chronic pain conditions [
15,
16].
The relationship between structural spinal changes and functional outcomes remains controversial [
17]. Traditional imaging-based definitions of spinal health have been challenged by high prevalence of structural abnormalities in asymptomatic individuals [
18-
21]. Disc degeneration (DD), Schmorl nodes (SN), and Modic change (MC) show variable associations with clinical symptoms and functional status [
18,
19,
21]. The growing dissociation between structural “abnormalities” and lived function suggests that judging interventions against an anatomy-pristine, fully painless comparator risks misclassifying clinically meaningful benefits as failures. In contrast, function-centered anchors, such as Oswestry Disability Index (ODI), gait speed, grip strength, and health utility (EuroQoL-5 dimensions [EQ-5D]), offer reproducible, patient-relevant metrics that align with what matters in practice. In this context, the ODI offers a pragmatic framework for operationalizing functional status in individuals with LBP [
22]. Within its widely used interpretive bands, scores around 20% or lower are typically categorized as “minimal disability,” indicating that pain produces little interference with usual ADL rather than complete symptom resolution. Building on this convention, we defined functional resilience in this study as the capacity to maintain minimal disability (ODI≤20%) despite the presence of chronic LBP and age-typical structural spinal changes on MRI. This function-centered definition does not aim to capture psychological resilience in a comprehensive sense, but rather to delineate a clinically observable phenotype of preserved day-to-day function in the setting of persistent symptoms and degenerative imaging findings.
We advance a pragmatic reframing, defining functional resilience as the capacity to maintain normal ADL despite LBP and common structural changes. We consider this to be a defining feature of spinal health. Understanding the characteristics of individuals who maintain function despite chronic LBP could inform development of interventions focused on functional preservation rather than pain reduction alone and contribute to more clinically relevant definitions of spinal health.
The primary aim of this study was to test the hypothesis that functional resilience exists among individuals with LBP to determine whether a substantial proportion maintains normal ADL despite the presence of structural MRI abnormalities and self-reported chronic pain.
MATERIALS AND METHODS
1. Study Population
This cross-sectional study was performed using a subcohort from the second visit of the Research on Osteoarthritis/Osteoporosis Against Disability (ROAD) study, which was initiated as a nationwide, prospective study of bone and joint diseases in population-based cohorts established in 3 communities with different characteristics (urban, mountainous, and coastal regions) in Japan. A detailed profile of the ROAD study has already been described elsewhere [
20,
21,
23].
The second visit of the ROAD study began in 2008 and was completed in 2010. All participants from the baseline study were invited to participate in the second visit. Additionally, inhabitants aged 60 years and older in the urban area and those aged 40 years and younger in the mountainous and coastal areas who were willing to participate were also included (both the mountainous and coastal areas were in Wakayama prefecture). From the 2,674 individuals who participated in the second visit of the ROAD study, we invited all 1,607 participants (547 men, 1,060 women) from the mountainous and coastal areas to the Wakayama Spine Study.
Of these 1,607 participants, 1,011 individuals (335 men, 676 women) provided written informed consent and attended the Wakayama Spine Study with MRI examinations. Among these participants, those with MRI-sensitive implanted devices (e.g., pacemakers) and other disqualifiers were excluded. A total of 145 participants were excluded from the analysis: 2 participants with pacemakers, 33 with unclear MRI findings, 9 who had undergone previous lumbar operations, and 101 with insufficient questionnaire responses. This resulted in 866 participants (285 men, 581 women) eligible for the present study [
20,
21].
These 866 participants were then classified according to the presence or absence of LBP. Among them, 347 participants (103 men, 244 women) reported LBP (1-month duration), while 519 participants (187 men, 332 women) reported no-LBP. The LBP group formed the primary focus of this study, as we aimed to investigate characteristics of individuals who maintain normal ADL despite chronic pain (
Fig. 1). Baseline characteristics of the entire cohort stratified by sex are summarized in
Table 1.
The study was conducted in accordance with the Declaration of Helsinki. The protocol was approved by the Ethics Committee of Wakayama Medical University, and written informed consent was obtained from all participants.
2. Assessment of Function and Pain
The ODI is a widely used questionnaire that assesses pain-related disability in people with LBP [
23]. It consists of 10 items addressing pain intensity, personal care, lifting, walking, sitting, standing, sleeping, sex life, social life, and traveling. Each item is scored from 0 to 5, with higher values representing greater disability. Functional impairment was defined using the ODI score. Participants were divided into 2 groups: ODI≤20% indicating maintained function and ODI≥21% indicating impairment. A threshold of 20% is widely used to represent the “minimal disability” category of the ODI, in which pain has little impact on ADL, whereas higher scores reflect increasing functional limitation. This operationalization aligns with the descriptive categories proposed by the developers and with prior epidemiological work using similar thresholds to distinguish individuals with preserved function from those with clinically relevant disability [
22].
Experienced orthopedists asked all participants the following questions regarding LBP: “Have you experienced LBP on most days during the past month, in addition to now?” Those who answered “yes” were defined as having LBP, as in previous studies [
20,
21]. Additionally, LBP intensity was assessed in all participants using a 100‑mm visual analogue scale (VAS), with 0 representing no pain and 100 representing the worst pain ever experienced.
All participants underwent MRI of the whole spine using a mobile MRI unit (Excelart 1.5 T, Toshiba, Japan) on the same day as the examination. The participants were supine during the MRI, and those with rounded backs used triangular pillows under their head and knees. The imaging protocol included sagittal T2-weighted fast spin echo (FSE) (repetition time [TR], 4,000 msec/echo; echo time [TE], 120 msec; field of view [FOV], 300 mm×320 mm), and axial T2-weighted FSE (TR, 4,000 msec/echo; TE, 120 msec; FOV, 180 mm×180 mm).
Sagittal T2-weighted images were used to assess the intervertebral space from C2–3 to L5–S1. DD grading was performed by an orthopedic surgeon (MT) who was blind to the background of the subjects. The degree of DD on MRI was classified into 5 grades based on Pfirrmann classification system [
24], with grades 4 and 5 indicating DD.
For grade 4, the signal intensity was intermediate to hypointense (black) compared to the cerebrospinal fluid, with an inhomogeneous structure. For grade 5, the signal intensity was hypointense (black) compared to the cerebrospinal fluid, with an inhomogeneous structure, and the disc space collapsed. Loss of signal intensity is significantly associated with the morphological level of DD and is also associated with both water and proteoglycan content in a disc [
20,
21]. Therefore, we used a grading based on signal intensity and disc height.
For the present analysis, lumbar DD was evaluated using the Pfirrmann grading system at each lumbar level from L1–2 to L5–S1, with grades 4 and 5 indicating degeneration. DD was defined as present if at least one lumbar disc from L1–2 to L5–S1 was graded as Pfirrmann grade 4 or 5.
For evaluating intraobserver variability, 100 randomly selected magnetic resonance images of the entire spine were rescored by the same observer (MT) more than 1 month after the first reading. To evaluate interobserver variability, 100 other magnetic resonance images were scored by 2 orthopedists (MT and RK) using the same classification. The intraobserver and interobserver variability for DD, as evaluated by kappa analysis, was 0.94 and 0.94, respectively.
MC was defined as diffuse areas of high signal change along the end plates, tending to be linear and always parallel to the vertebral end plates on sagittal T2-weighted images. However, discerning the type of MCs was not possible because of cost and time limitations of this large-scale study [
21]. Because the T1 sequence was not obtained, we considered Modic type I/II (T2 high signal intensity end plate change) to reflect the presence of MC and T2 iso-signal intensity and Modic type III (T2 low signal intensity end plate change) to reflect the absence of MC. To evaluate the intraobserver and interobserver variabilities, 2 orthopedic surgeons scored magnetic resonance images in the same manner. The intraobserver and interobserver variabilities of MC evaluated by kappa analysis were 0.86 and 0.82, respectively.
SN was characterized by a localized defect at the rostral, caudal, or both end plates, with a well-defined herniation pit in the vertebral body with or without a surrounding sclerotic rim (low signal on T2-weighted image) [
21]. Erosive defects in the end plate in degenerate segments were not considered as SN. To evaluate intraobserver and interobserver variabilities, 2 orthopedic surgeons scored magnetic resonance images in the same manner. The intraobserver and interobserver variabilities for SN evaluated by kappa analysis were 0.92 and 0.84, respectively.
The anthropometric measurements included height, weight, and body mass index (BMI). Measurement of BMI is more user-friendly and widely practiced. In this study, we decided to use BMI ≥25 kg/m
2 as an indicator of being overweight, based on the criteria of the Japan Society for the Study of Obesity [
25]. Alcohol consumption of ≤540 mL per day and current smoking status were assessed using a self-administered questionnaire. In addition, participants were asked whether their daily life or occupational activities involved lifting objects weighing 10 kg or more, and this was recorded as a binary indicator of habitual mechanical load.
Quality of life (QoL) was assessed using the 8-Item Short Form health survey (SF-8) (Japanese version), which provides physical composite score (PCS) and mental composite score (MCS). Health utility was measured with the EuroQoL-5 dimensions (EQ-5D) [
26].
All blood and urine samples were extracted between 9:00 AM and 3:00 PM. Some samples were extracted under fasting conditions. After centrifugation of the blood samples, sera were immediately placed in dry ice and transferred to a deep freezer within 24 hours. These samples were stored at -80°C until assayed. For the samples of participants in the baseline study, the following items were measured: blood counts, hemoglobin, hemoglobin A1c (HbA1c), blood sugar, total protein, aspartate aminotransferase, alanine aminotransferase, γ-glutamyl transpeptidase, high-density lipoprotein cholesterol (HDL-C), total cholesterol, triglycerides (TGs), blood urea nitrogen, uric acid, and creatinine. These analyses were performed at the same laboratory within 24 hours after extraction (Osaka Kessei Research Laboratories Inc., Japan).
Definitions of metabolic syndrome (MS) components were based mainly on the criteria of the Examination Committee of Criteria for MS in Japan [
27]. According to the consensus, an abdominal circumference ≥85 cm in men and ≥90 cm in women is a necessary condition for MS. Hypertension (HT) was diagnosed as systolic BP ≥130 mmHg and/or diastolic BP ≥85 mmHg; Dyslipidemia (DL), as serum TG level ≥150 mg/dL and/or serum HDL-C level <40 mg/dL; and impaired glucose tolerance (IGT), as fasting serum glucose level ≥100 mg/dL. Recently, the National Cholesterol Education Program’s Adult Treatment Panel III report proposed a new set of criteria to define MS without central obesity, as indicated by waist circumference, as the core feature [
28].
In addition, because not all blood samples were obtained under fasting conditions, we did not use participants’ data concerning serum levels of glucose and TGs because of their large variation depending on hours after eating. Instead, we used serum HDL-C level <40 mg/dL to indicate DL, and serum HbA1c level ≥5.5% to indicate IGT (the value for HbA1c [National Glycohemoglobin Standardization Program, NGSP]) (%) is estimated as an NGSP-equivalent value calculated by the formula HbA1c (%)=HbA1c (Japan Diabetes Society) (%)+0.4% [
29]. These are indices used in the National Health and Nutrition Survey in Japan, which were adopted as criteria for MS in this national screening based on the difficulty of collecting samples under fasting conditions [
30].
Participants were first classified into 4 groups according to the presence of LBP and disability status based on the ODI (LBP with disability, LBP without disability, no-LBP with disability, and no-LBP without disability). Baseline characteristics of the entire cohort stratified by sex are summarized in
Table 1. For comparisons across the 4 LBP/disability groups, continuous variables were compared using 1‑way analysis of variance (ANOVA), and categorical variables were compared using chi‑square tests. When ANOVA indicated a significant overall difference (p<0.05),
post hoc pairwise comparisons were performed using the Tukey-Kramer test to identify which groups differed while controlling for multiple comparisons.
To identify factors associated with maintained function among participants with LBP, we performed multivariable logistic regression analyses with disability status as the dependent variable (resilient: ODI ≤20% vs. impaired: ODI≥21%). Candidate predictors were selected from variables that showed significant associations with functional status in univariate analyses (
Table 2) and included age, sex, obesity (BMI ≥25 kg/m²), serum HDL‑C, diastolic blood pressure, LBP VAS, grip strength, and normal and maximal 2‑m walking times. All selected variables were entered simultaneously into a single multivariable model to estimate adjusted odds ratios (ORs) and 95% confidence intervals (CIs). Potential multicollinearity among predictors was assessed using variance inflation factors (VIFs) in the Fit Model platform of JMP; all VIF values were below 10, suggesting no serious multicollinearity. Analyses were conducted using JMP Student Edition (SAS Institute Japan, Japan), and a 2‑sided p-value <0.05 was considered statistically significant. The study was conducted in accordance with the Declaration of Helsinki.
RESULTS
1. Study Population Characteristics
This study analyzed data from 866 participants (285 men, 581 women; mean age, 66.4±13.5 years) from the Wakayama Spine Study. The cohort had a mean height of 156.4±9.4 cm, weight of 56.8±11.5 kg, and BMI of 23.3±3.6 kg/m². Overall, 10.9% of participants were smokers, and 30.8% consumed alcohol. Among the 347 participants with LBP, 220 (63.4%) maintained normal daily function (ODI≤20%), while 127 (36.6 %) were functionally impaired (ODI≥21%), confirming that a substantial subset maintains ADL despite pain (
Table 1).
Age strongly tracked disability status across the cohort (p<0.001). Within the LBP group, resilient participants were younger (65.0±11.9 years) than impaired participants (74.6±10.9 years). Body size differed across groups: height and weight showed significant differences (p<0.001 and p<0.05, respectively). BMI was highest in LBP with disability (24.1±3.5 kg/m²) and lowest in the no-LBP/nondisability group (22.9±3.5 kg/m², p<0.01). A male predominance in resilient LBP showed a trend (32.7% vs. 24.4% in disabled LBP; p=0.08). QoL-related measures, including the SF-8 PCS, MCS, and EQ‑5D, did not show significant differences across the 4 LBP/disability groups (
Table 2).
Tukey-Kramer
post‑hoc tests showed that BMI was significantly higher in participants with LBP and disability than in those without LBP and without disability, and that diastolic blood pressure was significantly higher in participants with LBP, irrespective of disability status, compared with those without LBP and without disability (
Supplementary Table 1). Serum HDL‑C was significantly lower in both disability groups than in the nondisabled group without LBP.
Group comparisons in spinal MRI demonstrated a notable dissociation between structural findings and functional status. DD was paradoxically more prevalent in resilient LBP participants (95.4%) than in those with disability (91.5%, p<0.05). SN were less prevalent in resilient LBP (32.9%) compared to those with disability (45.7%, p<0.001). MC showed a trend toward higher prevalence in those with disability (p=0.07), while LSS showed no group differences (p=0.79).
Crucially, multivariable logistic regression adjusted for age revealed that none of the MRI features were independently associated with functional status. DD showed no independent association (OR, 1.19; 95% CI, 0.46–3.08; p=0.7). Similarly, MC (OR, 0.80; 95% CI, 0.50–1.28; p=0.4), SN (OR, 0.83; 95% CI, 0.51–1.34; p=0.4) did not independently predict disability. These findings underscore that structural imaging burden, as captured by standard MRI phenotypes, is insufficient to determine real-world functional outcomes and thus poorly suited as a primary endpoint.
4. Metabolic and Biochemical Parameters
Across the 4 cohort groups, HDL-cholesterol was significantly higher in nondisabled strata (LBP without disability 63.7±15.3 vs. LBP with disability 60.6±14.0 mg/dL; no-LBP without disability 64.2±16.8 vs. no-LBP with disability 55.8±15.9 mg/dL; overall p<0.001). HbA1c levels did not differ significantly across groups, ranging from 5.2%±0.6% to 5.4%±0.8% (p=0.37). Diastolic blood pressure differed significantly across groups (p<0.001), with the LBP without disability group showing the highest values (77.3±10.9 mmHg) compared to LBP with disability (72.3±11.9 mmHg), no-LBP without disability (76.9±11.5 mmHg), and no-LBP with disability (73.4±10.9 mmHg). DL prevalence varied substantially across groups (p<0.001), ranging from 1.8% in LBP without disability to 14.3% in no-LBP with disability, compared to 5.4% in LBP with disability and 4.2% in no-LBP without disability.
In age-adjusted multivariate models, obesity (BMI ≥25 kg/m2) was the only metabolic factor independently associated with functional impairment (OR, 0.50; 95% CI, 0.30–0.84; p<0.01). Other metabolic components, including HT (OR, 0.98; p=0.93), DL (OR, 0.68; p=0.56), IGT (OR, 0.99; p=0.99), and the composite metabolic score (OR, 0.78; p=0.31), were not independently associated with functional status. Lifestyle factors, including smoking (OR, 0.58; p=0.13) and alcohol consumption (OR, 1.31; p=0.25), showed no independent association with maintained function.
5. Predictors of Functional Resilience Among Participants With LBP
Among participants with LBP, multivariable logistic regression using functional status (resilient: ODI ≤20% vs impaired: ODI ≥21%) as the dependent variable identified male sex and absence of obesity as independent predictors of maintained function (
Table 3). Male sex was associated with higher odds of resilience (OR, 1.76; 95% CI, 1.03–3.00; p<0.05), whereas obesity was associated with lower odds of resilience (OR, 0.50; 95% CI, 0.30–0.84; p<0.01). Smoking, alcohol consumption, lifting, metabolic components other than obesity, and MRI phenotypes (DD, MC, and SN) were not significantly associated with functional resilience after mutual adjustment.
To provide an integrated view of how clinical and lifestyle factors relate simultaneously to the presence of LBP and to functional status, we summarized their associations in a bubble chart (
Fig. 2). In this plot, the x-axis represents the odds of having LBP and the y-axis represents the odds of maintaining normal function, with each bubble corresponding to an individual factor and its size reflecting the magnitude of association; symbols indicate statistical significance (*p<0.05, **p<0.005). Factors such as obesity clustered in the quadrant indicating higher odds of LBP and lower odds of nondisability, whereas male sex and better physical performance measures shifted toward higher odds of nondisability. This visual representation reinforces the regression findings: obesity was negatively associated with functional resilience, whereas male sex and preserved physical performance were positively associated with maintaining function despite LBP.
Functionally resilient participants demonstrated superior physical performance across all measures compared to those with LBP-related disability. Grip strength was significantly higher in resilient participants (30.0±9.7 kg) compared to those with disability (23.5±8.2 kg, p<0.001). Walking speed also differed substantially: normal 2-m walking time was faster in resilient participants (5.16±1.38 seconds vs. 7.18±3.24 seconds, p<0.001), and maximal walking time was likewise faster (3.58±0.97 seconds vs. 5.00±2.15 seconds, p<0.001). These function-centered readouts (grip strength and gait speed) separated resilient from impaired LBP participants more clearly than imaging burden assessment. Among participants without LBP, those without disability also showed superior grip strength (29.5±9.1 kg) and faster normal gait speed (5.15±1.27 seconds) compared to those with disability (23.7±8.7 kg and 7.68±3.64 seconds, respectively; both p<0.001).
LBP VAS, grip strength, and normal and maximal 2‑m walking times also showed marked differences across the 4 groups (
Table 2). One‑way analysis of variance indicated highly significant group effects for all these variables (p<0.001). Tukey-Kramer
post hoc tests demonstrated that participants with LBP‑related disability had substantially higher pain scores than all other groups, including those with LBP but nondisability and those without LBP, and that participants with LBP but minimal disability had intermediate pain levels (
Supplementary Table 2). Similarly, both non‑disabled groups (with and without LBP) exhibited significantly greater grip strength and faster normal and maximal walking speeds than their respective disability counterparts, while performance was very similar between nondisabled participants with and without LBP.
In an exploratory subgroup analysis restricted to participants aged ≤70 years, the 4 LBP/disability groups showed broadly similar demographic and metabolic profiles (
Supplementary Table 1). Age, BMI, blood pressure, serum HDL‑C, HbA1c, obesity, HT, DL, and IGT did not differ significantly across groups, indicating that metabolic and anthropometric characteristics were less discriminative in this age‑restricted subset. In contrast, structural MRI abnormalities (DD, MC, and SN) remained highly prevalent and tended to cluster in participants with LBP, and symptom and function measures—including LBP VAS, SF‑8 PCS and MCS, EQ‑5D, and normal and maximal 2‑m walking times—continued to distinguish disabled from nondisabled participants regardless of LBP status. Grip strength showed a similar directional pattern, with somewhat lower values in disabled groups, although these differences were not statistically significant, consistent with the smaller sample size of the stratified analysis (
Supplementary Table 1).
DISCUSSION
In this population-based cohort, 63% of individuals reporting LBP nevertheless maintained minimal disability on the ODI. Within the conventional interpretive bands of the instrument, scores ≤20% correspond to the “minimal disability” category, in which pain has only a minor impact on usual daily activities rather than being completely absent. Our definition of functional resilience therefore reflects a pragmatic, function-centered phenotype, with preserved ADL despite chronic LBP and age-typical degenerative MRI changes, rather than a state of no pain or structurally normal spine. Although the ODI includes some pain-related items, it primarily measures the impact of pain on daily activities rather than pain intensity per se; our classification therefore reflects functional limitation in ADL rather than pain burden or global QoL. Our operational definition of functional resilience was based on minimal disability on the ODI and therefore focused on preserved ADL rather than on global QoL. In line with this distinction, QoL-related measures such as the SF‑8 PCS, MCS, and EQ‑5D did not show significant differences between groups in this cohort. These findings suggest that individuals who maintain functional independence in daily activities may still experience broader decrements in perceived health or well-being that are not fully captured by disability scores alone. Obesity was the only metabolic factor that consistently tracked with functional impairment, whereas canonical imaging findings (DD, MC, SN, and LSS) did not independently predict disability. Our findings show that individuals with LBP can function well despite structural abnormalities visible on MRI. This challenges the common practice of judging treatment success solely by the absence of pain and normal imaging findings, which is an unrealistic and clinically unhelpful standard.
Our results corroborate a durable theme in spine research: the weak correspondence between structural MRI abnormalities and clinical disability [
17-
19]. Consistent with Brinjikji et al. [
19], degenerative features were common and only loosely coupled to functional status. Notably, DD was more prevalent among functionally resilient LBP participants than among those with disability in unadjusted comparisons, and none of the evaluated MRI features independently explained functional classification after age adjustment. This strengthens the inference that anatomical burden, as captured by standard MRI phenotypes, is insufficient to determine real-world function and therefore poorly suited as a primary endpoint or as the defining attribute of “normality.”
Pain intensity is subjective, temporally labile, and sensitive to context and recall windows [
9-
13], “asymptomatic” status is neither stable nor fully objective across studies [
14,
15]. These limitations are particularly important given the weak correlation between pain intensity and functional impairment [
11,
12]. By contrast, function-centered measures (ODI, gait speed, grip strength, and health utility [EQ-5D]) demonstrated clear separation between resilient and impaired participants and map directly onto meaningful daily capacities. These are outcomes that matter to patients and clinicians. This finding is consistent with prior observations demonstrating the weak correspondence between structural MRI abnormalities and clinical disability [
18,
19]. Our findings therefore support replacing the conventional benchmark of symptom resolution and structural normality with a function-centered benchmark that is both clinically relevant and more reproducible, as advocated by recent work emphasizing outcome measures aligned with lived experience [
13].
Male sex showed a favorable association with maintained function after adjustment for age, which may reflect multiple interacting factors including differences in muscle mass and biomechanics [
31-
33], but also socio-behavioral responses to pain, coping strategies, and help-seeking behavior [
34,
35]. This warrants mechanistic study to distinguish biological from psychosocial contributors, without defaulting to reductive explanations based on muscle mass alone.
Obesity emerged as the only modifiable metabolic risk factor independently associated with functional impairment. Unlike age and sex, which are nonmodifiable, or structural MRI abnormalities, which showed no independent association with function, obesity represents a targetable intervention point. The mechanisms linking obesity to functional disability likely operate through multiple interconnected pathways. Obesity is a state of chronic systemic inflammation in which adipose tissue secretes pro-inflammatory cytokines and adipokines that promote inflammation and insulin resistance [
36,
37]. Our finding of elevated DL in disability groups is consistent with evidence that obesity-related lipid dysfunction directly impairs vascular function and may reduce disc perfusion [
38,
39].
Critically, obesity but not other metabolic components independently predicted disability, suggesting obesity-specific mechanisms rather than global cardiometabolic dysfunction. Individuals with obesity who maintain function may possess enhanced compensatory mechanisms, while those with disability may represent a threshold effect wherein physiological burdens exceed capacity. Weight reduction interventions may improve functional capacity and suggest that modifiable metabolic factors are primary determinants of functional outcomes. Therefore, weight management should be prioritized as a mechanism-based intervention strategy in LBP management to interrupt the obesity-inflammation-pain-disability cascade.
In this cohort, functional resilience was operationalized using the total ODI, with scores ≤20% corresponding to the conventional “minimal disability” band, in which pain has only minor impact on usual ADL rather than being completely absent [
11,
23]. Because the groups were defined based on ODI, differences in disability scores between resilient and impaired participants are expected by design. In this context, performance-based measures such as grip strength and gait speed should be interpreted as complementary descriptors of the ODI-defined phenotype rather than as independent validators of the classification. The observation that resilient participants showed better physical performance despite similar structural MRI abnormalities is consistent with prior work highlighting the limited correlation between degenerative imaging findings and disability [
19] and reinforces a shift toward function-centered assessment in LBP. At the same time, given that the ODI includes pain-related items and our psychological measures did not differ significantly between groups, our data should be viewed as capturing selected functional and metabolic components of a broader biopsychosocial framework, rather than providing a comprehensive test of all its psychological and neuromuscular dimensions.
Collectively, the associations we observed between functional status, obesity, and performance-based measures, in the absence of independent effects of MRI phenotypes, suggest that functional outcomes in LBP may involve multiple domains beyond structural pathology. However, in this study we directly assessed only selected metabolic and functional domains. These findings support the use of function-centered outcomes as clinically meaningful descriptors of everyday capacity, while underscoring that they operationalize functional resilience in activities of ADL, rather than resilience in the broader psychological sense. This conceptualization has important implications for both understanding the heterogeneity of functional outcomes in LBP and for designing interventions that target multiple modifiable domains [
40].
Several limitations warrant consideration. First, the cross-sectional design precludes causal inference regarding the direction of associations between obesity, metabolic factors, and functional status; reverse causality is possible, such that disability may promote weight gain through reduced physical activity rather than obesity causing disability. These associations can only be ascertained by a follow-up study clarifying the incidence and progression rates of functional decline in the same cohort.
Second, the participants included in the present study may not represent the general population since they were recruited from only 2 local areas in Wakayama prefecture. To confirm whether participants are representative of the Japanese population, we compared anthropometric measurements and frequencies of smoking and alcohol consumption between the general Japanese population and our study participants. While no significant differences in BMI were observed, participants in our study had healthier lifestyle profiles than those of the general Japanese population, suggesting potential healthy volunteer bias. This selection bias should be considered when generalizing the results of the present study.
Third, LBP was classified using a standardized binary question, which may have included participants with only minimal pain in the LBP(+) group; although concurrent VAS assessment partially mitigates this limitation, some misclassification of pain severity cannot be completely excluded. Additionally, we did not assess several potentially important confounders such as socioeconomic status, occupational mechanical stress, psychological variables (e.g., catastrophizing, fear-avoidance beliefs, pain coping strategies), baseline physical activity levels, or sleep quality, which may mediate associations between obesity and functional status. Fourth, the cohort was predominantly older and Japanese; generalizability to younger or more ethnically diverse populations requires investigation, as the relationship between symptoms and MRI phenotypes has been shown to vary across population cohorts [
18]. Fifth, one of the challenges of epidemiology is that population-based cohorts tend to reflect the clinical practice of primary care. In back pain, where the condition is chronic, patients become highly selected by the time they reach secondary care settings. Once clinical referral selection is applied, it may be that the ‘functionally resilient’ may be excluded and the patterns of pathology may change [
18]. In addition, we did not systematically assess key structural and systemic determinants of function in older adults, such as global spinal alignment on full-spine radiographs, osteoporotic vertebral fractures, or sarcopenia-related measures of muscle mass and quality. These unmeasured factors may substantially influence ADL and could contribute to residual variability in functional status beyond the MRI phenotypes, metabolic factors, and performance measures evaluated in this study. Finally, although we included a simple indicator of mechanical load (self-reported lifting of objects ≥10 kg in daily life or work), we did not obtain detailed quantitative information on occupational or leisure-time physical activity. As a result, activity-related exposure could not be fully characterized, and more comprehensive assessment of physical workload and habitual activity will be required in future studies of functional resilience.
A major strength of this study is its adequately powered cohort, which enabled robust detection of clinically meaningful differences between functionally resilient and nonresilient chronic LBP subgroups. Methodological rigor was further reinforced by excellent MRI reading reproducibility: intraobserver reliability was assessed by rescoring 100 randomly selected whole-spine MRIs after a >1-month interval by the same reader (MT), and interobserver reliability by independent scoring of another 100 MRIs by 2 orthopedists (MT, RK), both using identical criteria; kappa values were 0.82 to 0.94 for intraobserver and interobserver agreement. These near-perfect agreement metrics indicate highly reliable structural phenotyping and provide an important objective counterweight to self-reported variables (e.g., smoking and alcohol use), thereby strengthening internal validity and confidence in the imaging–function inferences.
CONCLUSION
The present study demonstrates that functional resilience in LBP is common and is not negated by structural MRI abnormalities. Male sex and nonobesity are independent predictors of maintained function, while standard MRI features show no independent association with functional status after age adjustment. These results support the use of function-centered outcome measures as primary endpoints in clinical practice, as they more accurately reflect real-world functional capacity than structural imaging. Obesity emerges as a modifiable risk factor, suggesting that weight management may improve functional resilience in individuals with LBP.
Supplementary Materials
NOTES
-
Conflict of Interest
The authors have nothing to disclose.
-
Funding/Support
This work was supported by H-25-Choujyu-007 (Director, NY), H25-Nanchitou (Men)-005 (Director, ST), and 201417014A (Director, NY) from the Ministry of Health, Labour and Welfare, a Grant-in-Aid for Scientifc Research (C 26861206) of JSPS KAKENHI grant. And Collaborating Research with NSF 08033011-00262 (Direc tor, NY) from the Ministry of Education, Culture, Sports, Science, and Technology in Japan. This study also was supported by grants from the Japan Osteoporosis Society (NY, HO), a grant from JA Kyosai Research Institute (HO), Japan Society for the Promotion of Science, Grants-in-Aid for Scientifc Research (KAKENHI) Research C (1 7 K 1 0 9 3 7) (MT), a Grant from the Japanese Orthopaedics and Traumatology Foundation, Inc (No. 287) (MT) and The Nakatomi Foundation (MT), The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
-
Acknowledgments
The authors wish to thank Mrs. Tamako Tsutsumi, Mrs. Kanami Maeda, and other members of the Public Office in Taiji Town for their assistance in locating and scheduling the participants for examinations.
-
Author Contribution
Conceptualization: MT, MR, JF, NY; Data curation: MT, YI, KN, HO, TI, NY, MY; Formal analysis: MT; Funding acquisition: NY; Methodology: MT; Project administration: HH, ST, HY; Writing – original draft: MT; Writing – review & editing: MR, HH, JF, YI, KN, RK, SM, HO, TI, ST, NY, MY, HY.
Fig. 1.Flowchart of participant selection and classification in the Wakayama Spine Study. MRI, magnetic resonance imaging.
Fig. 2.Bubble chart showing the associations between clinical and lifestyle factors and the odds of low back pain (x-axis) and nondisability (y-axis). *p<0.05. **p<0.005.
Table 1.Baseline characteristics of all participants stratified by sex
Table 1.
|
Characteristic |
Overall (n = 866) |
Male (n = 285) |
Female (n = 581) |
p-value for gender |
|
Demographic characteristics |
|
|
|
|
|
Age (yr) |
66.4 ± 13.5 |
67.2 ± 13.9 |
66.0 ± 13.4 |
0.200 |
|
Height (cm) |
156.4 ± 9.4 |
164.6 ± 7.2 |
151.5 ± 7.2 |
< 0.001 |
|
Weight (kg) |
56.8 ± 11.5 |
64.5 ± 11.6 |
53.0 ± 9.4 |
< 0.001 |
|
Body mass index (kg/m2) |
23.3 ± 3.6 |
23.6 ± 3.4 |
23.1 ± 3.7 |
< 0.050 |
|
Systolic BP (mmHg) |
139.8 ± 19.6 |
141.5 ± 18.5 |
138.9 ± 20.0 |
0.080 |
|
Diastolic BP (mmHg) |
76.0 ± 11.5 |
78.3 ± 12.4 |
74.9 ± 10.8 |
< 0.001 |
|
Serum level of HDL-C (mg/dL) |
62.9 ± 16.2 |
55.8 ± 15.0 |
66.4 ± 15.5 |
< 0.001 |
|
Serum level of HbA1c (%) |
5.7 ± 0.7 |
5.6 ± 0.6 |
5.7 ± 0.7 |
0.500 |
|
Prevalence of selected characteristics (%) |
|
|
|
|
|
Smoking habit |
10.9 |
11.1 |
10.9 |
1.000 |
|
Alcohol consumption |
30.8 |
32.5 |
30.0 |
0.500 |
|
Driving |
4.9 |
10.7 |
1.9 |
< 0.001 |
|
Lifting objects weighing ≥ 10 kg |
48.4 |
62.8 |
41.2 |
< 0.001 |
|
Prevalence of each metabolic abnormality (%) |
|
|
|
|
|
Obesity |
29.1 |
32.5 |
27.4 |
0.100 |
|
Hypertension |
75.2 |
79.2 |
73.3 |
0.060 |
|
Dyslipidemia |
4.6 |
10.0 |
1.9 |
< 0.001 |
|
Impaired glucose tolerance |
8.5 |
10.7 |
7.4 |
0.100 |
|
Symptoms |
|
|
|
|
|
Low back pain (%) |
40.0 |
35.6 |
42.2 |
0.060 |
|
Low back pain VAS |
25.5 ± 28.4 |
21.4 ± 25.6 |
27.6 ± 29.5 |
< 0.010 |
|
ODI score |
12.6 ± 14.4 |
11.2 ± 13.3 |
13.2 ± 14.9 |
0.050 |
|
SF-8 PCS |
45.5 ± 7.1 |
45.7 ± 7.1 |
45.4 ± 7.0 |
0.600 |
|
SF-8 MCS |
51.0 ± 6.6 |
51.0 ± 6.5 |
50.9 ± 6.7 |
0.800 |
|
EQ-5D |
0.84 ± 0.16 |
0.85 ± 0.17 |
0.85 ± 0.16 |
0.900 |
|
Spinal phenotype |
|
|
|
|
|
Presence of disc degeneration (%) |
90.1 |
88.2 |
91.0 |
0.200 |
|
Presence of Modic change (%) |
47.6 |
43.9 |
49.5 |
0.100 |
|
Presence of Schmorl node (%) |
33.6 |
28.7 |
36 |
< 0.050 |
|
Grip strength (kg) |
28.3 ± 9.5 |
37.5 ± 8.8 |
23.8 ± 5.8 |
< 0.001 |
|
Normal 2-m walking speed (sec) |
5.6 ± 2.2 |
5.7 ± 2.4 |
5.5 ± 1.5 |
0.100 |
|
Max 2-m walking speed (sec) |
3.9 ± 1.4 |
3.7 ± 1.1 |
4.0 ± 1.6 |
< 0.005 |
Table 2.Characteristics of participants between 4 groups
Table 2.
|
Characteristic |
Presence of LBP
|
Absence of LBP
|
p-value |
|
Disability (n = 127) |
Nondisability (n = 220) |
Disability (n = 69) |
Nondisability (n = 450) |
|
Demographic characteristics |
|
|
|
|
|
|
Sex, male:female |
31:96 |
72:148 |
24:45 |
158:292 |
0.080 |
|
Age (yr) |
74.6 ± 10.9 |
65.0 ± 11.9 |
75.9 ± 9.9 |
65.7 ± 12.0 |
< 0.001 |
|
Height (cm) |
151.9 ± 10.0 |
156.4 ± 9.1 |
152.4 ± 8.6 |
156.9 ± 9.1 |
< 0.001 |
|
Weight (kg) |
55.4 ± 12.1 |
56.6 ± 10.4 |
53.4 ± 9.1 |
57.4 ± 12.0 |
< 0.050 |
|
Body mass index (kg/m2) |
24.1 ± 3.5 |
23.4 ± 3.4 |
23.6 ± 3.9 |
22.9 ± 3.5 |
< 0.010 |
|
Systolic BP (mmHg) |
141.8 ± 19.6 |
138.3 ± 20.7 |
142.8 ± 18.2 |
139.4 ± 19.2 |
0.200 |
|
Diastolic BP (mmHg) |
72.3 ± 11.9 |
77.3 ± 10.9 |
73.4 ± 10.9 |
76.9 ± 11.5 |
< 0.001 |
|
Serum level of HDL-C (mg/dL) |
60.6 ± 14.0 |
63.7 ± 15.3 |
55.8 ± 15.9 |
64.2 ± 16.8 |
< 0.050 |
|
Serum level of HbA1c (%) |
5.3 ± 0.7 |
5.2 ± 0.6 |
5.4 ± 0.8 |
5.3 ± 0.8 |
0.400 |
|
Prevalence of selected characteristics (%) |
|
|
|
|
|
|
Smoking habit |
13.9 |
11.5 |
8.7 |
10.2 |
0.600 |
|
Alcohol consumption |
35.6 |
29.7 |
32.9 |
29.6 |
0.600 |
|
Driving |
3.5 |
3.4 |
6.3 |
5.8 |
0.500 |
|
Lifting objects weighing ≥ 10 kg |
54.8 |
47.7 |
63.2 |
44.7 |
< 0.050 |
|
Prevalence of each metabolic abnormality (%) |
|
|
|
|
|
|
Obesity |
38.0 |
27.4 |
34.3 |
26.6 |
0.060 |
|
Hypertension |
78.1 |
72.2 |
82.9 |
74.7 |
0.300 |
|
Dyslipidemia |
5.4 |
1.8 |
14.3 |
4.2 |
< 0.001 |
|
Impaired glucose tolerance |
9.3 |
7.3 |
11.6 |
8.4 |
0.700 |
|
Spinal phenotype (%) |
|
|
|
|
|
|
Presence of disc degeneration |
91.5 |
95.4 |
88.6 |
87.4 |
< 0.050 |
|
Presence of Modic change |
53.5 |
53 |
42.9 |
44.1 |
0.070 |
|
Presence of Schmorl node |
45.7 |
32.9 |
44.3 |
28.8 |
< 0.001 |
|
Symptoms |
|
|
|
|
|
|
Low back pain VAS |
57.6 ± 21.4 |
42.0 ± 21.0 |
20.2 ± 13.1 |
4.7 ± 13.1 |
< 0.001 |
|
SF-8 PCS |
46.0 ± 6.8 |
44.8 ± 7.9 |
45.5 ± 6.8 |
46.8 ± 6.5 |
0.200 |
|
SF-8 MCS |
50.4 ± 6.8 |
50.7 ± 6.8 |
51.6 ± 6.2 |
51.2 ± 6.6 |
0.500 |
|
EQ-5D |
0.87 ± 0.15 |
0.82 ± 0.17 |
0.88 ± 0.16 |
0.85 ± 0.16 |
0.060 |
|
Physical performance |
|
|
|
|
|
|
Grip strength (kg) |
23.5 ± 8.2 |
30.0 ± 9.7 |
23.7 ± 8.7 |
29.5 ± 9.1 |
< 0.001 |
|
Normal 2-m walking speed (sec) |
7.18 ± 3.24 |
5.16 ± 1.38 |
7.68 ± 3.64 |
5.15 ± 1.27 |
< 0.001 |
|
Max 2-m walking speed (sec) |
5.00 ± 2.15 |
3.58 ± 0.97 |
5.22 ± 2.03 |
3.55 ± 0.91 |
< 0.001 |
Table 3.Multivariable logistic regression analysis of factors associated with functional resilience (ODI≤20%) among participants with low back pain (reference: impaired function, ODI≥21%)
Table 3.
|
Variable |
Odds ratio |
95% CI |
p-value |
|
Sex, male vs. female |
1.76 |
1.03–3.00 |
< 0.05 |
|
Smoking habit |
0.58 |
0.28–1.18 |
0.10 |
|
Alcohol consumption |
1.31 |
0.83–2.08 |
0.30 |
|
Driving |
0.65 |
0.16–3.00 |
0.60 |
|
Lifting objects weighing ≥ 10 kg |
0.71 |
0.44–1.14 |
0.20 |
|
Metabolic factor |
|
|
|
|
Obesity |
0.50 |
0.30–0.84 |
< 0.01 |
|
Hypertension |
0.98 |
0.55–1.72 |
0.90 |
|
Dyslipidemia |
0.68 |
0.19–2.50 |
0.60 |
|
Impaired glucose tolerance |
0.99 |
0.44–2.26 |
1.00 |
|
Metabolic components |
0.78 |
0.49–1.26 |
0.30 |
|
MRI phenotype |
|
|
|
|
Presence of disc degeneration |
1.19 |
0.46–3.08 |
0.70 |
|
Presence of Modic change |
0.80 |
0.50–1.28 |
0.40 |
|
Presence of Schmorl node |
0.83 |
0.51–1.34 |
0.40 |
REFERENCES
- 1. Hartvigsen J, Hancock MJ, Kongsted A, et al. What low back pain is and why we need to pay attention. Lancet 2018;391:2356-67.
- 2. Hoy D, Bain C, Williams G, et al. A systematic review of the global prevalence of low back pain. Arthritis Rheum 2012;64:2028-37.
- 3. GBD 2016 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 328 diseases and injuries for 195 countries, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet 2017;390:1211-59.
- 4. GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020;396:1204-22.
- 5. Hoy D, Brooks P, Blyth F, et al. The Epidemiology of low back pain. Best Pract Res Clin Rheumatol 2010;24:769-81.
- 6. Katz JN. Lumbar disc disorders and low-back pain: socioeconomic factors and consequences. J Bone Joint Surg Am 2006;88 Suppl 2:21-4.
- 7. Deyo RA, Mirza SK, Turner JA, et al. Overtreating chronic back pain: time to back off? J Am Board Fam Med 2009;22:62-8.
- 8. Wong CK, Mak RY, Kwok TS, et al. Prevalence, incidence, and factors associated with non-specific chronic low back pain in community-dwelling older adults aged 60 years and older: a systematic review and meta-analysis. J Pain 2022;23:509-34.
- 9. Weiner DK, Haggerty CL, Kritchevsky SB, et al. How does low back pain impact physical function in independent, well-functioning older adults? Evidence from the Health ABC Cohort and implications for the future. Pain Med 2003;4:311-20.
- 10. Kovacs FM, Abraira V, Zamora J, et al. Correlation between pain, disability, and quality of life in patients with common low back pain. Spine (Phila Pa 1976) 2004;29:206-10.
- 11. Tonosu J, Takeshita K, Hara N, et al. The normative score and the cut-off value of the Oswestry Disability Index (ODI). Eur Spine J 2012;21:1596-602.
- 12. Kumar R, Alostaz M, Bansal A, et al. Describing functional low back pain disability in the American population: a crowdsourced evaluation of the Oswestry disability index. Spine J 2024;24:2165-71.
- 13. Foster NE, Anema JR, Cherkin D, et al. Prevention and treatment of low back pain: evidence, challenges, and promising directions. Lancet 2018;391:2368-83.
- 14. Zeppieri G, Bialosky J, George SZ. Importance of outcome domain for patients with musculoskeletal pain: characterizing subgroups and their response to treatment. Phys Ther 2020;100:829-45.
- 15. Sturgeon JA, Zautra AJ. Resilience: a new paradigm for adaptation to chronic pain. Curr Pain Headache Rep 2010;14:105-12.
- 16. Ramírez-Maestre C, de la Vega R, Sturgeon JA, et al. Editorial: resilience resources in chronic pain patients: the path to adaptation. Front Psychol 2019;10:2848.
- 17. Cheung KM, Karppinen J, Chan D, et al. Prevalence and pattern of lumbar magnetic resonance imaging changes in a population study of one thousand forty-three individuals. Spine (Phila Pa 1976) 2009;34:934-40.
- 18. Jamaludin A, Kadir T, Zisserman A, et al. ISSLS PRIZE in Clinical Science 2023: comparison of degenerative MRI features of the intervertebral disc between those with and without chronic low back pain. An exploratory study of two large female populations using automated annotation. Eur Spine J 2023;32:1504-16.
- 19. Brinjikji W, Luetmer PH, Comstock B, et al. Systematic literature review of imaging features of spinal degeneration in asymptomatic populations. AJNR Am J Neuroradiol 2015;36:811-6.
- 20. Teraguchi M, Yoshimura N, Hashizume H, et al. Prevalence and distribution of intervertebral disc degeneration over the entire spine in a population-based cohort: the Wakayama Spine Study. Osteoarthritis Cartilage 2014;22:104-10.
- 21. Teraguchi M, Yoshimura N, Hashizume H, et al. The association of combination of disc degeneration, end plate signal change, and Schmorl node with low back pain in a large population study: the Wakayama Spine Study. Spine J 2015;15:622-8.
- 22. eProvide: Mapi Research Trust. Oswestry Disability Index (ODI) [Internet]. Lyon (France): Mapi Research Trust; 2022. [2023 Sep 14]. Available from: https://eprovide.mapi-trust.org/instruments/oswestry-disability-index#descriptive_information.
- 23. Yoshimura N, Muraki S, Oka H, et al. Cohort profile: research on Osteoarthritis/Osteoporosis Against Disability study. Int J Epidemiol 2010;39:988-95.
- 24. Pfirrmann CW, Metzdorf A, Zanetti M, et al. Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine (Phila Pa 1976) 2001;26:1873-8.
- 25. WHO Western Pacific Region. The Asia-Pacific perspective: redefining obesity and its treatment. Sydney: Health Communications Australia; 2000.
- 26. Shiroiwa T, Fukuda T, Ikeda S, et al. Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D. Qual Life Res 2016;25:707-19.
- 27. Matsuzawa Y. Metabolic syndrome--definition and diagnostic criteria in Japan. J Atheroscler Thromb 2005;12:301.
- 28. Grundy SM, Cleeman JI, Daniels SR, et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005;112:2735-52.
- 29. Japan Diabetes Society Committee on the Diagnostic Criteria of Diabetes Mellitus. Report of the Committee on the classification and diagnostic criteria of diabetes mellitus. J Jpn Diabetes Soc 2010;53:450-67.
- 30. Ministry of Health, Labour and Welfare. The outline of the results of National Health and Nutrition Survey. Tokyo (Japan): Ministry of Health, Labour and Welfare; 2008.
- 31. Tseng LA, Delmonico MJ, Visser M, et al. Body composition explains sex differential in physical performance among older adults. J Gerontol A Biol Sci Med Sci 2014;69:93-100.
- 32. Shiri R, Karppinen J, Leino-Arjas P, et al. The association between obesity and low back pain: a meta-analysis. Am J Epidemiol 2010;171:135-54.
- 33. Stalling I, Gruber M, Bammann K. Sex differences in physical functioning among older adults: cross-sectional results from the OUTDOOR ACTIVE study. BMC Public Health 2024;24:1766.
- 34. Cornally N, McCarthy G. Help-seeking behaviour for the treatment of chronic pain. Br J Community Nurs 2011;16:90-8.
- 35. Sanchez J, Brooks J, Fitzgerald S, et al. Pain coping profiles in workers’ compensation clients with chronic musculoskeletal pain. Aust J Rehabil Couns 2015;21:104-23.
- 36. Reilly SM, Saltiel AR. Adapting to obesity with adipose tissue inflammation. Nat Rev Endocrinol 2017;13:633-43.
- 37. Saltiel AR, Olefsky JM. Inflammatory mechanisms linking obesity and metabolic disease. J Clin Invest 2017;127:1-4.
- 38. Li M, Qian M, Kyler K, et al. Adipose tissue-endothelial cell interactions in obesity-induced endothelial dysfunction. Front Cardiovasc Med 2021;8:681581.
- 39. Yuan L, Huang Z, Han W, et al. The impact of dyslipidemia on lumbar intervertebral disc degeneration and vertebral endplate modic changes: a cross-sectional study of 1035 citizens in China. BMC Public Health 2023;23:1302.
- 40. Wasser JG, Vasilopoulos T, Zdziarski LA, et al. Exercise benefits for chronic low back pain in overweight and obese individuals. PM R 2017;9:181-92.
, Marinko Rade2,3
