Long-Term Prognosis and Risk Factors for Low Back Pain-Related Disorders in the General Population: A 7-Year Follow-Up of the Wakayama Spine Study

Abstract

Introduction

Low back pain (LBP) is a leading cause of disability worldwide, particularly in aging populations. While the Oswestry Disability Index (ODI) is widely used to assess LBP-related disability, few studies have evaluated its long-term trajectory and predictive factors in general populations.

Methods

This 7-year longitudinal study included 553 community-dwelling adults (mean age 66.3 years) from the Wakayama Spine Study, a population-based sub-cohort of the Research on Osteoarthritis/Osteoporosis against Disability (ROAD) study. Participants completed whole-spine magnetic resonance imaging and responded to the ODI questionnaire at baseline and follow-up. Disability levels were classified as mild (0%-20%), moderate (21%-40%), or severe (41%-60%). Longitudinal transitions in disability categories were analyzed descriptively. Multiple linear regression was used to identify predictors of ODI deterioration.

Results

The mean ODI score significantly increased from 9.6±11.5 at baseline to 12.2±14.2 after 7 years (p<0.001), although the change did not reach clinical significance. Among participants initially classified as mildly disabled (n=468), 88.0% remained stable, while 12.0% worsened. Of those with moderate disability (n=73), 35.6% improved, 44.0% remained unchanged, and 20.5% worsened. No participant with severe disability (n=40) improved to mild. Female sex, older age, higher body mass index, and vertebral fractures (semiquantitative grade ≥2) were significant predictors of worsening disability (p<0.05). Higher baseline ODI was inversely associated with deterioration.

Conclusions

In this population-based cohort, LBP-related disability modestly worsened over 7 years. Older adults, women, individuals with obesity, and those with vertebral fractures were at greatest risk. These findings support early intervention and screening strategies to prevent disability progression in at-risk populations.

Introduction

Low back pain (LBP) is recognized as one of the leading causes of physical disability and diminished quality of life worldwide, with its impact becoming increasingly significant due to the growing burden among older adults in aging societies^1,2). LBP is a prevalent health concern among both adults and the elderly, substantially impairing activities of daily living and physical function^3,4). The etiology of LBP is multifactorial, involving not only physical factors such as age, sex, body mass index (BMI), lifestyle habits (including smoking and alcohol consumption), decreased bone mineral density (BMD), vertebral fractures, and spinal canal stenosis, but also psychosocial factors such as anxiety, stress, and fear-avoidance beliefs and behaviors⁵⁾.

The Oswestry Disability Index (ODI) is a widely utilized instrument for assessing the degree of functional impairment associated with LBP^6,7). It quantitatively evaluates the impact of LBP on daily activities and has been validated as a valuable tool in both clinical and epidemiological research^6,7). However, evidence regarding longitudinal changes in ODI scores and the identification of their predictive factors within the general population remains limited.

The present study aimed to elucidate the longitudinal changes in ODI scores and identify their predictive factors over 7 years among 1,011 community-dwelling residents in Wakayama Prefecture, Japan. We placed particular emphasis on examining the influence of sex, age, BMI, pre-existing vertebral fractures, spinal canal stenosis, and osteoporosis (OP) on changes in ODI scores. The findings of this study will contribute to a deeper understanding of risk factors associated with LBP and offer critical insights for the development of future prevention and intervention strategies.

Materials and Methods

Ethics approval

The study followed the Declaration of Helsinki and was ethically approved. All participants gave written informed consent. All procedures complied with institutional and national ethical standards.

Participants

This study was conducted as part of the Research on Osteoarthritis/Osteoporosis against Disability (ROAD) study, a population-based cohort on musculoskeletal health in Japan^8,9). Protocols are detailed elsewhere¹⁰⁾. This analysis used data from Hidakagawa (a mountainous area) and Taiji (a coastal area) in Wakayama Prefecture.

The second survey of the ROAD study in these regions was conducted between 2008 and 2010. Inclusion required mobility, questionnaire completion, and consent. Excluded: tumors, infections, inflammation, prior fusion, magnetic resonance imaging (MRI)-incompatible devices, and pregnancy. Overall, 1,063 residents were invited for whole-spine MRI using a mobile unit (Excelart 1.5T, Toshiba, Tokyo). Among them, 52 declined MRI scanning, and two were excluded due to contraindications, resulting in a baseline cohort of 1,009 participants (335 men and 674 women; mean age, 66.3 years; age range, 2-97 years). The fourth follow-up survey of the ROAD study, conducted between 2015 and 2016, enabled longitudinal assessment after approximately 7 years. We attempted to follow all 1,009 baseline participants. A total of 348 individuals were lost to follow-up due to death (n=95), illness (n=35), relocation (n=17), time constraints (n=45), absence (n=3), refusal to participate (n=33), or unknown reasons (n=120). Among the 661 remaining participants, 98 individuals had missing data for the main outcome measure. Ultimately, 553 participants with complete datasets were included in the final analysis (Fig. 1).

Figure 1.

Flow diagram of participant selection and follow-up in the Research on Osteoarthritis/Osteoporosis against Disability (ROAD) study.

Participants were recruited from two communities in Wakayama Prefecture as part of the second survey of the ROAD study conducted between 2008 and 2010. Of the 1,063 individuals who consented to participate, 52 declined magnetic resonance imaging (MRI), and 2 were excluded due to pacemakers, resulting in a baseline cohort of 1,009 participants. After 7 years of follow-up (2015-2016), 348 participants were lost to follow-up. Among the 661 remaining individuals, 98 were excluded due to missing data for the main outcome, and a total of 553 participants were included in the final analysis.

Anthropometric measurements

Anthropometric measurements, including height and weight, were measured in all participants. BMI=weight (kg)/height (m²).

Radiographic assessment

Plain radiographs of the cervical, thoracic, and lumbar spine in the anteroposterior and lateral views, and bilateral knees in the anteroposterior view with weight-bearing and foot-map positioning were obtained.

Magnetic resonance imaging

A mobile MRI unit (Achieva 1.5 T; Philips Medical Systems, Best, The Netherlands) was used, and whole-spine MRI was performed for all participants on the same day as the examination. The participants were supine during the MRI, and those with rounded backs used triangular pillows under their heads and knees. The imaging protocol included sagittal T2-weighted fast spin echo imaging (repetition time, 3,000 ms/echo; echo time, 120 ms; and field of view, 270×270 mm) and axial T2-weighted fast spin echo imaging (repetition time, 2,100 ms/echo; echo time, 100 ms; and field of view, 180×180 mm). Sagittal images covered the spine; axial images spanned T12/L1-L5/S1.

Spinal lateral X-rays and MRI assessments were performed at baseline and repeated after 7 years. However, in this study, we analyzed baseline imaging data to investigate the association between pre-existing spinal conditions and subsequent changes in ODI scores over the follow-up period.

Evaluation of intervertebral disc degeneration

Sagittal T2 images assessed discs C2/3-L5/S1. C2/3 to C7/T1, T1/2 to T12/L1, and L1/2 to L5/S1 were defined as the cervical region, thoracic region, and lumbar region, respectively. Disc degeneration (DD) grading was performed by an orthopedist (MT) who was blind to the background of the subjects. The degree of DD on MRI was classified into five grades based on Pfirrmann's classification system¹¹⁾, with grades 4 and 5 indicating DD. The signal intensity for grade 4 was intermediate to hypointense to the cerebrospinal fluid (CSF) (dark gray), while the structure is inhomogeneous. Grade 5: hypointense to cerebrospinal fluid CSF (black), inhomogeneous. For grade 5, the disc is hypointense to CSF (black), inhomogeneous, and shows significant disc space collapse, indicating severe DD. It has been reported that loss of signal intensity is significantly associated with the morphological level of the DD and is also associated with both the water and proteoglycan content in a disc¹²⁾. The Pfirrmann classification for DD was based on both signal intensity and disc height, with grades 4 and 5 indicating advanced degeneration¹³⁾. Intra- and interobserver Kappa values for DD were both 0.94.

Evaluation of radiographic lumbar spinal stenosis

Radiographic lumbar stenosis was qualitatively assessed post-examination. An experienced orthopedic surgeon (YI) without knowledge of the participants' symptom status examined the images, which were provided on films. Axial images were used to rate central stenosis. We used the classification included in a general guideline^8,14) in which mild stenosis was defined as narrowing of one-third of the normal area or less, moderate stenosis as narrowing of between one- and two-thirds, and severe stenosis as narrowing of more than two-thirds of the area. The intra- and interobserver reliability for central stenosis grading was high, with Kappa values of 0.82 and 0.77, respectively.

Evaluation of cervical cord compression

Cervical cord compression (CCC) was graded from C2/C3 to C7/T1 by an experienced orthopedic surgeon (KN). Grading was defined as follows: grade 0=no compression of the spinal cord, with the subarachnoid space remaining intact; grade 1=no compression of the spinal cord with subarachnoid space absent; grade 2=compression of less than one-third of the spinal cord; grade 3=compression of more than one-third but less than two-thirds of the spinal cord; and grade 4=compression of more than two-thirds of the spinal cord. CCC was defined as grade ≥2 at the most affected level⁹⁾. Intraobserver variability was assessed by rescoring 100 cervical MRIs (KN) after ≥1 month. Interobserver variability was assessed using 100 MRIs scored by another orthopedic surgeon (HH). The intraobserver and interobserver variabilities for CCC evaluated by Kappa analysis were 0.78 and 0.72, respectively.

Evaluation of vertebral fractures

Pre-existing vertebral fractures in the thoracolumbar region were evaluated on sagittal MRI images using the Genant semiquantitative (SQ) method¹⁵⁾. This classification assesses vertebral deformities by visually estimating the reduction in vertebral body height and categorizing fractures into three grades: grade 1 (mild, 20%-25% height reduction), grade 2 (moderate, approximately 25%-40% reduction), and grade 3 (severe, more than 40% reduction). Both anterior, middle, and posterior heights were taken into consideration to distinguish true fractures from non-fracture deformities. The total score (0-9) was used as an index by summing the respective grades of the Th11-L1 vertebrae. The prevalence of vertebral fractures was reported to be highest in Th11-L1 lesions¹⁶⁾. To evaluate the intraobserver variability, 100 randomly selected MRIs of the entire spine were rescored by the same observer (MT) more than 1 month after the first reading. Furthermore, to evaluate interobserver variability, 100 other MRIs were scored by two experienced orthopedic surgeons (MT and RK) using the same SQ method. Kappa values for vertebral fracture scoring were 0.87 (intra) and 0.84 (inter).

Evaluation of diffuse idiopathic skeletal hyperostosis of the spine

Diffuse idiopathic skeletal hyperostosis (DISH) was diagnosed according to the following criteria, defined by Resnick and Niwayama¹⁷⁾: (1) flowing ossification along the lateral aspect of at least four contiguous vertebral bodies, (2) relative preservation of intervertebral disc height in the involved segments, and (3) absence of epiphyseal joint bony enclosing and sacroiliac joint erosion. Due to poor visibility of C7/T1-T3/4, ‘whole-spine’ referred to C0/1-C6/7, Th4/5-Th12/L1, and L1/2-L5/S1¹⁸⁾.

BMD examination

Lumbar spine and proximal femur BMD values were determined using dual-energy X-ray absorptiometry (DXA; Hologic Discovery C; Hologic, Waltham, MA). The precision of the DXA system was monitored using a BMD phantom, which showed measurement variability within ±1.5% across all measurements. To control observer variability, the same physician (NY) examined all participants. Intraobserver variability of DXA using the Lunar DPX system (GE Healthcare, formerly Lunar Corporation, Madison, WI, USA) had been previously assessed by the same physician (NY) in another study¹⁹⁾, with a coefficient of variance for L2-4 in vitro of 0.35%. In vivo CVs for L2-4, proximal femur, Ward's triangle, and trochanter were 0.61%-0.90%, 1.02%-2.57%, 1.97%-5.45%, and 1.77%-4.17%, respectively. In the present study, OP was defined using the World Health Organization criteria, which classify OP as a BMD T-score 2.5 standard deviations or more below the young adult mean (peak bone mass)²⁰⁾. In Japan, the mean (±standard deviation [SD]) L2-4 BMD in young adults, as measured using Hologic DXA, is 1.011 (±0.119) g/cm^{2 21)}, and lumbar spine OP was defined as L2-4 BMD <0.714 g/cm². For the femoral neck, the mean (±SD) BMDs in young adult males and females are 0.863 (±0.127) and 0.787 (±0.109) g/cm², respectively²¹⁾, and OP was defined as a BMD of <0.546 and <0.515 g/cm² for men and women, respectively.

Questionnaire about the presence of LBP

LBP was assessed by interview: pain lasting for more than 24 hours within the past month and localized between the lower margin of the 12th rib and the inferior gluteal folds²²⁾. During the interview, participants were shown an illustration depicting the defined anatomical region of LBP and were asked, ‘Have you experienced LBP on most days (and continuously on at least one day) in the past month, in addition to any current pain?’ Those who responded ‘yes’ were classified as having LBP.

ODI

The ODI is a widely used and highly reliable instrument for evaluating disability in individuals with LBP. It comprises 10 sections that assess various aspects of daily living, including pain intensity, walking, sitting, and social life. Each section is scored 0-5; higher scores indicate greater disability. The total score is adjusted for any unanswered sections and expressed as a percentage (0%=no disability; 100%=maximum disability)^6,7).

In the present study, due to a high rate of non-responses to the sex life section (Section 8), likely influenced by cultural factors, this item was excluded for all participants^23,24). Accordingly, the ODI score was calculated based on the remaining nine sections, and the percent disability was derived from the sum of the available item scores. The severity of disability was categorized according to the classification proposed by Fairbank et al. as follows: 1 (minimal): 0%-20%; 2 (moderate): 21%-40%; 3 (severe): 41%-60%; and 4 (crippled): 61%-80%^6,7). No participants in this study had an ODI score exceeding 80%^6,7).

Statistical analysis

Descriptive statistics are shown as mean±SD or frequency (%) unless noted. Differences in proportions were examined using the chi-square test, or Fisher's exact test when the expected cell size was <5. For continuous variables, Student's t-test was used following confirmation of normality via the Shapiro-Wilk test. Changes in the prevalence of self-reported LBP and ODI % disability over the 7 years were first examined. Subsequently, predictors of changes in ODI % disability were analyzed using multiple linear regression. In this study, ‘changes in ODI % disability’ were defined as the difference between the post-ODI score and the pre-ODI score. This approach was chosen to preserve the full granularity of change in the ODI, avoiding information loss from dichotomization, and to quantify associations between continuous baseline variables (e.g., age, sex, BMI, vertebral fractures) and the magnitude of ODI change. However, we acknowledge that a logistic regression analysis based on a clinically meaningful threshold, such as minimal clinically important difference (MCID), could provide complementary insights. To address this, we performed an additional logistic regression analysis categorizing participants into two groups (ODI deterioration ≥11% vs <11%) and included the results in the supplementary materials. The explanatory variables included sex, age, BMI, smoking status, alcohol consumption, presence of OP (determined at the femoral neck), presence of vertebral fractures of grade 2 or higher, lumbar DD (defined as the sum of Pfirrmann grades from L1/2 to L5/S)²⁵⁾, CCC, lumbar spinal canal stenosis, DISH, and baseline ODI score. All statistical analyses were performed using JMP Pro version 17 (SAS Institute Japan, Tokyo, Japan), with the level of statistical significance set at p<0.05.

Results

The baseline characteristics of the participants are summarized in Table 1. This table shows the distributions of demographic data, lifestyle factors, and spinal characteristics at the beginning of the study. At baseline, the prevalence of self-reported LBP was 40.5%, which increased to 45.4% after 7 years. The mean ODI score increased significantly from 9.6±11.5 (mean±standard deviation) at baseline to 12.2±14.2 at the 7-year follow-up (mean change: +2.6 points, p<0.001, paired t-test). However, this mean change did not reach the threshold for clinical significance.

Table 1.

Characteristics of the 553 Participants at the Baseline.

Variables	Men	Women
Number of participants	187	366
Age (years)	65.1±11.7	62.7±11.0
Age strata (number)
<50	18 (9.6%)	51 (13.9%)
50-59	48 (25.7%)	97 (26.5%)
60-69	46 (24.6%)	111 (30.3%)
70-79	52 (27.8%)	85 (23.2%)
≥80	23 (12.3%)	22 (6.0%)
Height (cm)	165.3±6.6	153.0±6.2
Weight (kg)	65.6±10.8	54.2±9.1
Body mass index (kg/m2)	24.0±3.2	23.1±3.6
Alcohol drinking habit	61 (32.6%)	122 (33.3%)
Smoking habit	25 (13.4%)	43 (11.8%)
Low back pain	70 (37.4%)	154 (42.1%)
ODI score (%)	8.8±10.6	10.0±12.0
Lumbar DD score (L1/2-L5/S1)	18.0±2.3	17.9±2.4
Radiographic LSS (grade 3)	52 (27.8%)	94 (25.8%)
CCC (≥grade 2)	49 (26.2%)	60 (16.4%)
Vertebral fracture index (Th11-L1)	2.3±1.5	1.7±1.6
Vertebral osteoporosis	2 (1.1%)	41 (11.2%)
DISH	32 (17.1%)	13 (3.6%)

Data are presented as mean±standard deviation or as number (%).

CCC: cervical cord compression; DD: disc degeneration; DISH: diffuse idiopathic skeletal hyperostosis; LSS: lumbar spinal stenosis; ODI: Oswestry disability index

In contrast, analysis of individual participants revealed that 95 participants (17.2%) experienced deterioration in ODI scores exceeding the MCID threshold of 11 points⁷⁾, while 40 participants (7.2%) demonstrated improvement beyond the same threshold.

According to the categorical classification of ODI disability levels at baseline, 84.6% of participants were classified as having mild disability, 13.2% as moderate, and 2.2% as severe. After 7 years, the number of participants classified as mild decreased by 30, whereas those classified as moderate, severe, or crippled increased (Table 2).

Table 2.

Presence of Low Back Pain and ODI at Baseline and 7 Years Later.

Items	Baseline	7 years later
Presence of low back pain	224/553 (40.5%)	250/551 (45.4%)
ODI % disability (mean±SD)	9.6±11.5	12.2±14.2
ODI category
Mild	468/553 (84.6%)	438/553 (79.2%)
Moderate	73/553 (13.2%)	81/553 (14.6%)
Severe	12/553 (2.2%)	33/553 (6.0%)
Crippled	0/553 (0%)	1/553 (0.2%)

This table summarizes the prevalence of self-reported low back pain and ODI scores among participants at baseline and after a 7-year follow-up. It includes mean ODI scores with SD, and categorical distributions of ODI disability levels (mild, moderate, severe, and crippled). In addition, 95 participants (17.2%) experienced deterioration in ODI scores exceeding the MCID threshold of 11 points, while 40 participants (7.2%) demonstrated improvement beyond the same threshold.

MCID: minimal clinically important difference; ODI: Oswestry disability index; SD: standard deviation

A longitudinal analysis of disability transitions revealed that among the 468 participants initially classified as mild, 412 (88.0%) remained in the same category, while 12% showed progression. Of the 73 participants with moderate disability at baseline, 35.6% improved to mild, 43.8% remained unchanged, and 20.6% worsened. Importantly, none of the participants with severe disability at baseline improved to the mild category after 7 years; 75% remained in the severe category (Fig. 2).

Figure 2.

Longitudinal transitions in disability status over 7 years.

This flowchart illustrates the 7-year transitions in Oswestry Disability Index (ODI) severity among older adults in the general population. Among participants classified as having a mild disability at baseline (n=468), 412 (88.0%) remained mild and 56 (12.0%) progressed to moderate or severe. Of those with moderate disability (n=73), 26 participants (35.6%) improved to mild, 32 (43.8%) remained moderate, and 15 (20.6%) progressed to severe. Among participants with severe disability at baseline (n=40), none (0%) improved to mild, and 9 (75.0%) remained severe. Values are presented as number (percentage). Statistical analysis was descriptive; no hypothesis testing was performed for transitions in this figure.

To identify predictors of ODI deterioration, multiple linear regression analysis was performed. The analysis revealed that female sex (standardized β=0.15, p=0.0003), older age (β=0.23, p<0.0001), higher BMI (β=0.13, p=0.0027), and the vertebral fracture index at the thoracolumbar region (β=0.12, p=0.0115) were significant risk factors for increased disability. Conversely, a higher baseline ODI score was found to be a significant protective factor (β=-0.32, p<0.0001), indicating that individuals with greater initial disability were less likely to experience further deterioration during follow-up (Table 3).

Table 3.

Risk Factors for ODI Deterioration.

Variables	Standardized β	P-value	VIF
Sex (female)	0.15	0.0003	1.16
Age	0.23	<0.0001	1.59
BMI	0.13	0.0027	1.11
Smoking habit	0	0.9621	1.07
Drinking habit	0.05	0.2036	1.06
Lumbar osteoporosis (L2-L4)	-0.01	0.8014	1.13
Vertebral fracture index (Th11-L1)	0.12	0.0115	1.13
Lumbar DD score (L1/2-L5/S)	0.03	0.5165	1.32
Cervical cord compression (≥grade 2)	-0.01	0.8348	1.07
Lumbar spinal stenosis (grade 3)	0.04	0.3613	1.11
DISH (presence)	-0.03	0.5165	1.13
ODI (%) at baseline	-0.32	<0.0001	1.16

Vertebral fracture index was calculated from the sum of Genant grade among T11-L1, Lumbar DD score was calculated from the sum of Pfirrmann grade among L1/2-L5/S. Boldface indicates statistically significant variables and values.

BMI: body mass index; DD: disc degeneration; DISH: diffuse idiopathic skeletal hyperostosis; VIF: variance inflation factor

Discussion

This longitudinal population-based study investigated temporal changes in LBP-related disability and its determinants among 553 community-dwelling residents of Wakayama Prefecture over 7 years. Self-reported LBP increased from 40.5% to 45.4%, with mean ODI rising from 9.6 to 12.2. These trends reflect the gradual progression of LBP with aging. While the statistical significance of the increase in ODI scores was evident, the mean change of +2.6 points did not surpass the commonly accepted MCID for ODI, which typically ranges from 10 to 12 points^7,26,27). This suggests that although a deterioration in function was measurable at the population level, it may not reflect a perceptible decline in individual patient-reported outcomes. This discrepancy underscores the need for clinical context when interpreting disability data.

Beyond mean changes, our categorical analysis of ODI severity levels offers a more nuanced perspective on the trajectories of disability. The vast majority (88.0%) of individuals initially classified as having mild disability remained stable over time, while a subset demonstrated progression to moderate or severe levels. Participants with severe baseline disability showed minimal improvement, highlighting its persistence. These findings suggest that timely intervention, ideally before functional impairment becomes entrenched, may be critical to preventing irreversible decline.

Our multivariable regression analysis identified four key risk factors independently associated with worsening ODI scores over time: female sex, old age, high BMI, and the presence of vertebral fractures (SQ grade ≥2). These variables are established contributors to LBP and disability¹⁾. The association between female sex and ODI deterioration may be attributed to a higher risk of vertebral fractures and DD due to lower BMD and hormonal changes, along with psychosocial factors such as depression and caregiving burden. However, we recognize that our primary analysis used a linear regression model to explore continuous changes in ODI. To complement this, we conducted an additional logistic regression analysis with ODI deterioration defined by a clinically meaningful threshold (MCID, 11% increase). The logistic model demonstrated that while certain factors were associated with deterioration in univariate analyses, only age remained significant in multivariate analysis (see Supplementary Tables 1 and 2). This underscores the importance of considering both continuous and categorical approaches when evaluating longitudinal outcomes. These findings are echoed in the work of Leboeuf-Yde et al.²⁸⁾, who reported stronger associations between obesity and LBP in women than in men in a twin study cohort. From a mechanistic standpoint, Samartzis et al.²⁹⁾ have postulated that obesity may exert deleterious effects on the spine both mechanically―by increasing axial load―and biochemically, through adipokine-mediated inflammation that facilitates nociceptive ingrowth into the intervertebral disc²⁹⁾.

The role of pre-existing vertebral fractures in disability progression is also consistent with prior studies employing the Genant SQ grading system. Fracture-related deformities impair posture, mobility, and quality of life³⁰⁾. Our findings reinforce the imperative of OP prevention and fracture risk assessment as components of spinal health maintenance, particularly in aging populations.

Paradoxically, a higher baseline ODI score was associated with less deterioration over time―a finding that diverges from conventional expectations and is seldom reported. One plausible interpretation is that individuals with greater initial disability may have been more likely to pursue targeted interventions or modify their lifestyle during the follow-up period. Alternatively, this phenomenon may reflect statistical regression to the mean³¹⁾. Being observational, this study cannot fully separate behavioral effects from statistical artifacts. Nevertheless, this counterintuitive finding opens avenues for future investigation into behavioral adaptation and self-management among individuals with chronic disability.

From a public health perspective, these results underscore the urgency of preventive strategies targeting individuals at greatest risk of disability progression―namely, older adults, women, those with obesity, and individuals with vertebral fractures. Managing these risk factors early may reduce the population burden of LBP-related disability.

However, several limitations merit consideration. First, the study sample was derived exclusively from a single prefecture in Japan, and caution is warranted in extrapolating these findings to other geographic or cultural contexts. Epidemiological patterns of LBP may vary with urbanization, healthcare access, and social norms. Second, self-reported data may introduce bias in assessing pain and function. Subjectivity in symptom reporting can result in underestimation or exaggeration of disability. Third, psychosocial variables such as psychological distress, anxiety, and depression―which have been shown to be important determinants of chronic pain―were not systematically evaluated in this study³²⁾. Future studies should incorporate these factors for a more comprehensive disability model. Fourth, while information on medication use and physical therapy was collected at both baseline and 7-year follow-up, these data were not included in the present analysis. This omission may introduce residual confounding, and future studies should comprehensively account for the influence of treatment interventions on changes in ODI scores.

In summary, our study delineates the natural history of LBP-related disability in the general population over 7 years and identifies several key risk and protective factors. The findings support early, tailored interventions and proactive spinal health maintenance.

In conclusion, this 7-year longitudinal study among community-dwelling individuals identified female sex, older age, higher BMI, and pre-existing vertebral fractures as independent risk factors for worsening disability associated with LBP, as measured by the ODI. These findings highlight the need for targeted preventive strategies and early interventions in high-risk populations to curb the progression of disability and maintain functional independence in aging societies. The results contribute important evidence to the understanding of LBP trajectories in the general population and may inform future clinical and public health efforts aimed at reducing its burden.

Disclaimer: Hiroshi Hashizume is one of the Editors of Spine Surgery and Related Research and on the journal's Editorial Committee. He was not involved in the editorial evaluation or decision to accept this article for publication at all.

Conflicts of Interest: The authors declare that there are no relevant conflicts of interest.

Author Contributions: Conceptualization: Naomi Iwane (NI), Shizumasa Murata (SM), and Hiroshi Hashizume (HH). Data curation: Hiroshi Iwasaki (HI), Masanari Takami (MTa), Shunji Tsutsui, Keiji Nagata (KN), Yuyu Ishimoto (YI), Masatoshi Teraguchi (MT), and Toshiko Iidaka (TI). Formal analysis: SM, HH, and Hiroyuki Oka (HO). Investigation: NI, SM, and HH. Methodology: SM, HH, and Hiroshi Yamada (HY). Project administration: HH, Noriko Yoshimura (NY), Sakae Tanaka (ST), and HY. Resources: HO. Software: HO. Supervision: ST, NY, and HY. Validation: Kanae Mure, SM, HH, and HO. Visualization: SM. Writing - original draft: NI and SM. Writing - review & editing: HH.

Hiroshi Yamada and Noriko Yoshimura contributed equally to this work.

Ethical Approval: This study was approved by the Research Ethics Committee of Wakayama Medical University (No. G120) and The University of Tokyo (No. 1326). All participants provided written informed consent for the inclusion and publication of their data. All procedures were performed following the ethical standards of the responsible committee on human experimentation (institutional and national) and the Declaration of Helsinki.

Informed Consent: All patients signed informed consent regarding publishing their data and photographs.

Data Availability: The datasets generated and/or analyzed during the current study are available only to approved co-investigators of the ROAD Study, following the data-sharing policy jointly defined by the Research Ethics Committees of The University of Tokyo.

Requests for data access may be considered on a case-by-case basis, subject to approval by the study steering committee. The data are not publicly available due to ethical and privacy restrictions.

Funding Statement

This study was supported by Grant-in-Aid funding from the Ministry of Health, Labour, and Welfare 21FA0601 (Director, HY), and Scientific Research grants 26462249, 21K09306 (HH). The study was also supported by Scientific Research grants B19H03895, B26293139, B23390172, and B20390182, and Challenging Exploratory Research grants 18K18447, 15K15219, and 24659317 (NY). The study was partly supported by grants from the Japan Agency for Medical Research and Development (AMED) JP22dk0110047 (KM). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.