Abstract
Background
Low back pain(LBP) has become a major public health problem worldwide. Assessing the attributable contribution of modifiable risk factors in the burden of LBP and its epidemiological trends is crucial for designing effective interventions and public health strategies.
Methods
This study comprehensively analyzes the global burden of LBP attributable to modifiable risk factors from 1990 to 2021 and projects trends to 2041, using data from the GBD study, the CHARLS, and GWAS. Age‑sex analyses, cohort‑based non‑linear relationship analyses, and two‑sample Mendelian randomization were also performed.
Results
Based on the GBD analytical framework, our results identify environmental/occupational risks, smoking, and high body mass index as the top contributors to LBP-attributable burden. Projections indicate a rising burden of LBP attributable to high BMI globally, particularly in middle and low SDI regions, while the burden from smoking is declining. Age‑sex analyses reveal the highest burden among middle‑aged and older adults, with a higher prevalence in females. Analyses within the CHARLS cohort demonstrated a significant non‑linear relationship between BMI and LBP risk. Two‑sample Mendelian randomization further provided evidence consistent with a causal effect of high BMI on LBP.
Conclusion
These results underscore the urgent need for targeted public health strategies addressing these modifiable risks across different regions, ages, and sexes to mitigate the growing global burden of LBP.
Keywords: low back pain, high body mass index, disability-adjusted life years, global burden, risk factors
Introduction
Low back pain (LBP) is among the commonest musculoskeletal issues, and it has a marked impact on how patients experience their daily quality of life. Reportedly, the lifetime prevalence of LBP ranks among the highest worldwide, ranging approximately from 50% to 90%, and thus represents a serious global public health challenge.1–3 LBP is commonly defined as pain localised to the area between the lower margin of the twelfth rib and the gluteal folds, which may or may not be accompanied by radiating pain to one or both lower limbs. This definition may also encompass pain in the buttock region.4 It is estimated that ~80% of adults experience at least one acute LBP episode during their lifetime, consistent with global lifetime prevalence estimates. Point prevalence (current LBP at a given time) in adults is typically 10%–30%, and incidence (new cases per year) is 5%–15% per year. The clinical course of LBP is highly variable, and estimates of persistent or recurrent pain at one year vary substantially by case definition, population, and clinical setting. In community and primary-care cohorts, approximately 40%–60% of individuals with mechanical low back pain report persistent symptoms or frequent recurrences one year after onsetPMC. For new-onset lumbar radiculopathy, 15%–40% develop chronic or recurrent pain.5–7 In the United States, a 2013 study by Hong et al estimated that among patients with chronic low back pain, the total annual direct medical costs per patient were $8,615, with nonoperative treatments (including medications, physical therapy, and injections) accounting for 26.3% of the total annual direct costs. For context, the denominator of this 26.3% figure is the total annual direct medical costs per patient, which includes all medical services (inpatient, outpatient, emergency, pharmacy, etc.) related to low back pain; the numerator comprises the portion attributable to nonoperative management. In the United Kingdom, a 2006 analysis by Maniadakis and Gray (using a bottom-up, prevalence-based approach, expressed in 2004–2005 GBP) estimated the total annual cost of low back pain at £2.8 billion, comprising direct healthcare costs (£1,393 million) and indirect costs (£1,448 million). In Australia, a 2012 study by Schofield et al estimated that the total annual cost of spinal disorders exceeded AUD 4,800 million, incorporating lost income, taxation losses, and government support payments. While these estimates are derived from different years, currencies, and methodological frameworks, precluding direct cross-country comparisons, they collectively underscore the substantial economic burden imposed by LBP on healthcare systems and society.8–10 Therefore, given the rising economic and social burden caused by LBP, individuals, families, society, and healthcare systems face substantial challenges. To tackle this issue, all relevant parties must work in collaboration to lessen LBP’s impact and develop long-term strategies that reduce the burden borne by individuals and medical systems.
LBP is a multifactorial musculoskeletal disorder shaped by biological, psychological, social, and occupational determinants. While these factors collectively contribute to LBP onset and persistence. Although multiple biopsychosocial factors contribute to the pathogenesis and chronicity of low back pain (LBP), this study focuses specifically on three modifiable risk factors—high BMI, smoking, and environmental/occupational ergonomic exposures—for the following reasons.11 First, these three factors are consistently identified as the dominant contributors to the global LBP burden in the GBD framework, which systematically quantifies attributable disease burden across 88 risk factors. Second, unlike non-modifiable factors such as genetics or age, these factors are amenable to population-level interventions, making them highly relevant for public health policy. Third, previous GBD studies have analyzed these risk factors separately, but a direct comparative projection of their future burden across different socioeconomic regions remains lacking. Therefore, while we acknowledge the multifactorial nature of LBP—including psychosocial, comorbid, and genetic influences—our study prioritizes these three modifiable drivers to fill a critical evidence gap and inform targeted intervention strategies.12,13 Based on a comprehensive global framework, this study adopts an overall perspective to better predict the epidemiological distribution, worldwide, of modifiable LBP risk factors over the next two decades, thereby filling a gap in the literature. The study provides trends in the burden of risk factors at global and regional levels and conducts stratified analyses by sex and age; using Autoregressive Integrated Moving Average (ARIMA) models to project future trends to 2041 and joinpoint regression to analyze temporal trends from 1990 to 2021, thereby elucidating the distinct patterns of risk-factor burden. What distinguishes this study from previous GBD-based reports is not merely an update of estimates or the addition of new datasets. Rather, the novelty lies in methodological triangulation. We integrate three distinct lines of evidence to converge on a focused public health question: (1) GBD forecasting to compare the diverging future trajectories of smoking versus high BMI; (2) independent cross-sectional analysis of the CHARLS cohort to characterize the non-linear, individual-level relationship between BMI and LBP; and (3) two-sample Mendelian randomization to provide evidence consistent with a causal effect of high BMI on LBP. This triangulation approach allows us to move beyond descriptive burden estimation and offer a more robust, multi-evidence justification for prioritizing specific modifiable risks—particularly high BMI—in future global LBP prevention strategies. These findings help improve population-level health and offer policymakers and healthcare system planners profound insights to prioritize key risk factors by specific ages, sexes, and regions, thereby addressing the root causes of LBP.
Data and Methods
Study Data
GBD Database
The 2021 GBD study offers extensive epidemiological insights, encompassing 371 diseases and injuries alongside 88 risk factors, and spanning 204 countries and territories over the three-decade period from 1990 to 2021.14,15 Within this research, LBP was defined in alignment with the ICD coding framework, with specific inclusion of pertinent ICD-9 and ICD-10 code sets. Following this definition, keyword searches identified the following three modifiable risk factors: environmental or occupational hazards, tobacco use, and high BMI. Additionally, DALYs served as the core measurement to quantify the overall burden of both fatal and non-fatal conditions on population health; this metric combines YLLs due to premature mortality and YLDs due to disease-related impairment. For LBP, mortality contributes negligibly to total DALYs; the burden is overwhelmingly driven by disability. In the GBD 2021 dataset, YLLs account for <1% of LBP-related DALYs globally, while YLDs constitute >99% of the total burden. This distinction is critical to avoid misinterpretation: LBP is not a fatal condition, and its population impact stems almost entirely from chronic pain, functional limitation, reduced quality of life, and work disability. The present analysis focuses on disability-dominated burden attributable to modifiable risk factors, reflecting the true public health relevance of LBP as a leading cause of non-fatal disability worldwide.
CHARLS Database
The China Health and Retirement Longitudinal Study (CHARLS) is a longitudinal survey targeting Chinese adults aged 45 years and older, with biennial follow-up assessments initiated in 2011. It employs a multi-stage sampling design stratified by urban and rural administrative divisions, supplemented by probability-proportional-to-size (PPS) sampling.16 In the present study, cross-sectional analysis was performed using only the 2015 wave of CHARLS data to minimize attrition bias inherent to longitudinal studies, while this wave also demonstrates the highest completeness of key variables. In CHARLS, low back pain (LBP) was ascertained via self-report. This national cross-sectional analysis included 21,097 participants at baseline. After excluding individuals with missing data on key exposure and outcome variables (BMI, LBP status, age, and sex) and those with BMI values outside 10–80 kg/m2, the final analytical sample comprised 15,777 participants. In this sample, other covariates (education, alcohol use, hypertension, diabetes) had missing rates below 10% (hypertension: 17.2%, diabetes: 17.6%, others <1%).17
Statistical Analysis
This study integrates three complementary data sources to address distinct but related questions: (1) GBD quantifies population-level burden, trends, and projections of LBP attributable to modifiable risk factors; (2) CHARLS examines individual-level non-linear association between BMI and LBP in a Chinese cohort; (3) two-sample MR tests for evidence consistent with a causal effect of high BMI on LBP. Each analysis has specific inferential boundaries: GBD provides population estimates but does not imply causality; CHARLS adjusts for measured confounders but is cross-sectional and China-specific; MR supports causal inference but does not capture non-linearity or predict future population burden. Convergence across layers strengthens overall conclusions.
GBD Analysis
Risk Factor
Prior research has established that LBP is associated with multiple risk factors, such as environmental or occupational hazards, smoking, and high BMI.11 Building on this finding, the present study first quantified the proportion of DALYs attributed to each of these LBP-related risk factors. Subsequently, it projected how the DALYs proportion corresponding to each risk factor would change over the next two decades.
ARIMA Model
This study employed the ARIMA model to forecast crude and age-standardized DALYs attributable to LBP risk factors globally and across five SDI regions over the next two decades. The ARIMA model, denoted as ARIMA(p, d, q), incorporates autoregressive (p), differencing (d), and moving average (q) components. To construct the optimal model, the study used the auto.arima() function based on the AIC value—this function is suitable for univariate time series data and can screen different ARIMA models and determine the optimal scheme according to the set constraint order.18,19 At the same time, the prediction performance of the ARIMA model was evaluated by indicators such as RMSE, MAE, MAPE and CV_MAPE. All fitted ARIMA models satisfied the stationarity and white noise assumptions (ADF p > 0.05, Ljung–Box p > 0.05). In addition, RMSE values below 0.5 indicated reasonable forecast accuracy for most models, overall low RMSE, MAE and MAPE together with acceptable CV-MAPE further verified stable predictive performance of selected models (Supplementary Table 1).
Descriptive Analysis
Leveraging the GBD study’s defined summary data on geographical regions, a world map was generated via the ggplot2 and sf packages in R software (version 4.4.1) to visually illustrate the geographical distribution of LBP disease burden, and then analyze its global distribution characteristics and regional differences. At the same time, a stratified analysis based on the population was carried out to explore the distribution pattern of LBP in different demographic groups, specifically divided by age, gender, and specific subgroups. The above analysis results not only reflect the scale of LBP disease burden that the global healthcare system will face in the future, but also reveal the impact of population aging and population growth on LBP disease burden through age-standardized rates. Specifically, the age-standardized rate refers to the weighted average of disease rates across distinct age groups, where the weight corresponds to the share of the population in each respective age group within the standard population. All data provide a 95% UI, which provides statistical guarantee for the reliability of the research results.20
Join-Point Regression Model
The Joinpoint Regression Model, initially introduced by Kim in 2000,21 is a analytical approach that builds a piecewise regression framework according to the temporal patterns of disease distribution. Its core function lies in conducting trend fitting and optimization for data points within each segmented interval. In the present study, this Joinpoint regression method was adopted to analyze the APC and AAPC of LBP disease burden spanning from 1990 to 2021, with the aim of quantifying its temporal trends. Specifically, APC reflects the average annual rate of change in LBP disease burden over a defined time segment, while AAPC represents this rate across the entire duration of the study. The statistical criteria for judging trends were set as follows: A statistically significant upward trend in LBP disease burden during a given period is confirmed if the 95% CI of APC is entirely above 0; conversely, a statistically significant downward trend is indicated when the 95% CI of APC is fully below 0. If the 95% CI of APC includes 0, it means there is no statistically significant change in LBP disease burden over that period.
CHARLS Analysis
The study was based on the 2015 wave of the CHARLS database. The key exposure of interest was BMI, derived from measured weight and height (kg/m2) and treated as both a continuous and a categorical variable in the analyses. The model incorporated the following covariates: age (as a continuous variable), sex, educational attainment (dichotomized into low [≤ junior high school] and high [≥ senior high school]), smoking status (smoker [current or former] or non-smoker), alcohol use (drinker [current or former] or non-drinker), and a history of chronic diseases (hypertension or diabetes, based on self-reported physician diagnoses). These variables were collected through standardized household interviews. The outcome, LBP status, was ascertained through participant self-report (“yes”/“no”). Multiple imputation (MI) was performed using MICE (5 imputed datasets) to handle missing covariates (education, smoking, alcohol, hypertension, diabetes). As a sensitivity analysis, we repeated all descriptive analyses using complete-case analysis (CCA), ie, excluding any participant with missing covariate data. Key estimates (LBP prevalence, mean BMI, obesity proportion, and BMI category P-value) were compared between MI and CCA.
In this study, continuous variables are reported as mean ± standard deviation, while categorical variables are presented as counts with corresponding percentages. Given the results of the normality test, continuous variables were analyzed via the Wilcoxon rank-sum test, and categorical variables were examined using the chi-square test. Baseline characteristics were compared between the LBP group and the non-LBP group. To investigate the non-linear relationship between BMI and the risk of low back pain, a RCS was fitted, adjusting for confounders such as age, gender, education level, smoking, alcohol consumption, hypertension, and diabetes. The number of knots was automatically selected between 3 and 7 based on the AIC minimization principle, and Rubin’s rules were applied to the multiple imputation results for combination. We assessed how BMI correlates with low back pain risk using two tests: the overall association test (P overall) and the non-linear association test (P nonlinear). No population sampling weights were applied to the cross-sectional analyses of CHARLS data in this study. We did not conduct formal correction for multiple comparisons across these exploratory analyses. Such correction approaches would impose overly strict statistical thresholds in observational epidemiological research, which may obscure clinically meaningful associations between BMI and LBP.
MR Analysis
To further explore the causal link between BMI and LBP, we additionally performed a two-sample MR analysis. For this analysis, LBP outcome data were derived from the fin-b-M13_LOWBACKPAIN dataset, while summary-level GWAS data for the exposure—body mass index (BMI, identifier: ukb-b-19953)—were obtained from the Open GWAS project (https://gwas.mrcieu.ac.uk/). Instrumental variables (IVs) were selected using genome-wide significance threshold (p < 5×10−8) and LD clumping (r2 < 0.001, window = 10,000 kb). The F-statistic for each SNP was calculated, and the overall mean F-statistic was > 10, indicating no weak instrument bias. In terms of instrumental variable (IV) selection, 184 SNPs from the BMI GWAS were utilized as IVs. The IVW method served as the primary statistical strategy for conducting the MR analysis.
All statistical computations in the present study were carried out using R software (version 4.4.1), and a P-value of less than 0.05 was defined as the threshold for statistical significance. In addition to statistical significance testing, we interpreted effect sizes to enhance the clinical and practical relevance of the findings. For group comparisons, approximate reference values of 0.10, 0.30, and 0.70 for Cohen’s d or Hedges’ g were used as tentative interpretive guidelines for small, medium, and large effects in pain research, alongside appropriate sample size planning.22
Results
Risk Factors
In 2021, there were significant differences at the global and regional levels in the proportion of LBP DALYs attributable to specific risk factors. Globally, the contributions to LBP-related DALYs mainly came from smoking (12.6%), high BMI (11.9%), and environmental or occupational risks (22.2%) (Figure 1A). By sex, the primary risk factors for males were smoking and environmental or occupational risks, whereas for females they were environmental or occupational risks and high BMI (Figure 1B–C).
Figure 1.
Percentage of LBP-related DALYs attributable to three major risk factors stratified by GBD regions, SDI categories and sex. ((A) All individuals in 2021; (B) Males in 2021; (C) Females in 2021; (D) All individuals in 2041; (E) Males in 2041; (F) Females in 2041).
To further assess the principal risk factors contributing to LBP over the next 20 years, we projected that environmental or occupational risks would remain predominant in 2041 (26.2%, 95% UI: 18.4–33.9), an increase of 4 percentage points from the current level; this was followed by high BMI (15.5%, 95% UI: 14.4–16.5), an increase of 3.6 percentage points. In contrast, the burden of LBP attributable to smoking showed a declining trend, a pattern that was consistent for both males and females (Figure 1D–F).
Regionally, the disease burden attributable to high BMI is on the rise, while environmental or occupational risks are increasing markedly in low-middle SDI, low SDI and economically disadvantaged regions, but decreasing in high SDI, high-middle SDI and economically developed regions.
Global, 5SDI and National Level
In 2021, the global DALYs of LBP attributable to smoking, high BMI, and environmental or occupational risks were approximately 8.82 million, 8.36 million, and 15.57 million, respectively. From 1990 to 2021, the global age-standardized DALYs rate (ASDR) for LBP attributable to high BMI increased by 39.1% (95% UI: 31.5–45)., while that attributable to smoking decreased by 33.4% (95% UI: 31.7–35.3), and environmental/occupational risks decreased by 11.3% (95% UI: 7.6–14.7) (Figure 2A–C and Supplementary Table 2). In high SDI regions, smoking and high BMI were the primary risk factors for ASDRs; whereas environmental or occupational risks were the main contributors in low SDI regions (Figure 2D–F and Supplementary Table 2).
Figure 2.
DALYs burden of LBP caused by various risk factors in 2021 globally and in 5 SDI regions. ((A) Number of High BMI. (B) Number of Smoking. (C) Number of environmental or occupational risks. (D) ASDR of High BMI. (E) ASDR of Smoking. (F) ASDR of environmental or occupational risks).
At the country level, ASDRs of LBP attributable to high BMI were highest in Hungary, Montenegro, and the United States, at 280, 253, and 240, respectively; LBP ASDRs attributable to smoking were highest in Montenegro and Serbia, at 360 and 337, respectively; and LBP ASDRs attributable to environmental or occupational risks were highest in Mozambique, Nepal, and Albania, at 486, 461, and 452, respectively (Figure 3A–C).
Figure 3.
ASDR Disease Burden of LBP in 204 Countries in 2021. ((A) ASDR of High BMI; (B). ASDR of Smoking; (C) ASDR of environmental or occupational risks).
Forecast Analysis Results
Globally, the DALYs of LBP attributable to smoking, high BMI, and environmental or occupational risks are projected to reach approximately 10 million, 12.89 million, and 18.98 million, respectively, by 2041 (Figure 4A–C). While overall DALYs are projected to continuously increase, the ASDRs attributable to smoking and environmental or occupational risks are projected to decrease; however, the ASDR attributable to high BMI is projected to continue to increase. Across different SDI regions, the smoking-related ASDR is projected to decrease in all regions; the ASDR attributable to high BMI is projected to increase in all regions; and the ASDR attributable to environmental or occupational risks is projected to vary across regions, increasing in the high-middle and low SDI regions and decreasing in other regions (Figure 4D–F).
Figure 4.
Predict trends in low back pain DALYs globally and for each of the five SDI regions by contributing risk factors. ((A) Number of High BMI. (B) Number of Smoking. (C) Number of environmental or occupational risks. (D) ASDR of High BMI. (E) ASDR of Smoking. (F) ASDR of environmental or occupational risks).
Age and Sex Patterns
Age–sex analysis indicates that the DALY burden of LBP attributable to the three risk factors is concentrated mainly among middle-aged and older adults. Specifically: for DALYs counts attributable to smoking globally, males peak at ages 50–54 and then decline with increasing age; females peak at ages 55–59 and then gradually decline. Regarding DALYs rates, males reach their peak at ages 60–69 and females at ages 55–65, after which rates decline with advancing age (Figure 5A and B). For DALYs counts attributable to high BMI, global males peak at ages 50–54 and then decline with age; females peak at ages 50–59 and subsequently decline. For DALYs rates, both males and females peak at ages 70–74, followed by reductions as age increases (Figure 5C and D). For DALYs counts attributable to environmental or occupational risks, males peak between ages 45–55 and then decline with age; females peak at ages 50–54 and then gradually decline. Regarding DALYs rates, males peak at ages 55–59 and females at ages 50–54, with rates decreasing thereafter as age increases (Figure 5E and F).
Figure 5.
Age-gender trends in LBP burden in 2021. ((A) Rate of Smoking; (B) Number of Smoking; (C) Rate of High BMI; (D) Number of High BMI; (E) Rate of environmental or occupational risks; (F) Number of environmental or occupational risks).
Results of Joinpoint Regression Analysis
To further investigate the trends in LBP burden attributable to three risk factors, we performed Joinpoint regression analyses. The results indicate that from 1990 to 2021, DALYs for LBP attributable to high BMI showed an overall increasing trend globally and across the five SDI regions. Specifically, the AAPCs for DALYs due to high BMI for the globe and the five SDI regions were 1.07 (1.06 to 1.09), 1.00 (0.98 to 1.01), 0.74 (0.72 to 0.76), 2.07 (2.06 to 2.09), 2.32 (2.31 to 2.34), and 1.93 (1.91 to 1.95), respectively. In terms of period-specific changes, the global APC rose from 0.43% in 1990–1994 to a peak of 1.32% in 2000–2004, with minor fluctuations thereafter. The high SDI region exhibited an increasing trend before 2005 and then gradually declined; by contrast, the middle- and low-SDI regions maintained relatively high or persistently increasing APCs throughout the study period, indicating that the disease burden in these regions continues to worsen (Figure 6A and Supplementary Table 3).
Figure 6.
Joinpoint regression analysis results of DALYs. ((A) High BMI; (B) Smoking; (C) environmental or occupational risks. *, APC value with statistical significance at P<0.05).
Regarding the impact of smoking on LBP burden trends, the results show that from 1990 to 2021 the age-standardized DALYs rates for LBP attributable to smoking generally declined globally and across SDI regions. Specifically, the AAPCs for DALYs due to smoking for the globe and the five SDI regions were −1.31 (−1.35 to −1.27), −1.13 (−1.15 to −1.11), −0.83 (−0.85 to −0.81), −1.03 (−1.06 to −0.99), −1.02 (−1.05 to −0.98), and −1.32 (−1.37 to −1.27), respectively. For period-specific changes, the global APC was −1.66% during 1990–1993 and fluctuated in subsequent periods—for example, the APC was −1.38% during 2011–2021. The high SDI region had an APC of −1.52% in 1990–1993, which gradually slowed thereafter, reaching −1.04% in 2014–2021; meanwhile, some periods in middle- and low-SDI regions still showed substantial declines—for instance, the low SDI region had an APC of −1.64% in 2015–2019—suggesting that although smoking-related LBP burden has decreased in these regions, significant challenges persist during certain intervals (Figure 6B and Supplementary Table 4).
For environmental or occupational risks and their impact on LBP burden trends, the results indicate that from 1990 to 2021 the DALYs rates for LBP attributable to these risks generally declined globally and across SDI regions. Specifically, the AAPCs for LBP-related DALYs for the globe and the five SDI regions were −0.68 (−0.69 to −0.66), −0.39 (−0.40 to −0.38), −0.95 (−0.96 to −0.94), −0.67 (−0.68 to −0.66), −0.44 (−0.46 to −0.42), and −1.11 (−1.12 to −1.10), respectively. In period-specific changes, the global APC was −1.33% in 1990–1993 and showed fluctuations thereafter, for example −0.30% in 2019–2021. The high SDI region had an APC of −0.96% in 1990–1994, with a subsequent slowdown in the trend and even a slight increase during 2010–2018 (APC = 0.09%), before declining again in 2018–2021 (−0.40%). Some periods in middle- and low-SDI regions still displayed marked downward trends—for example, the low SDI region had an APC of −1.61% in 2015–2019—indicating that although the overall burden is decreasing, certain regions still faced substantial challenges during specific periods (Figure 6C and Supplementary Table 5).
Results of Analysis in the CHARLS Population
The crowd screening flowchart can be found in Figure 7. This study included 15,777 participants, of whom 47% were male and 53% were female. The prevalence of LBP in this cohort was 20.6%. Univariate analysis revealed that patients in the LBP group were older, had a higher proportion of females, and a higher proportion of individuals with lower education levels compared to the non-LBP group. When BMI was used as a categorical variable, there was a statistically significant difference between the groups (p = 0.020), specifically, a higher proportion of underweight and overweight individuals in the LBP group. Furthermore, the proportions of smokers, alcohol consumers, and patients with hypertension and diabetes were significantly higher in the LBP group (p < 0.001), see Table 1.
Figure 7.
Flow chart of the study population selection process.
Table 1.
Basic Characteristics of the Population in CHARLS
| Total Sample (n=15,777) | No LBP (N = 12,525) | LBP (N = 3252) | P-Value | |
|---|---|---|---|---|
| Age (years,mean±SD) | 59.6 ± 10.3 | 59.4 ± 10.3 | 60.4 ± 10.2 | <0.001 |
| Sex | <0.001 | |||
| Male | 7,343 (47%) | 6,277 (50%) | 1,066 (33%) | |
| Female | 8,434 (53%) | 6,248 (50%) | 2,186 (67%) | |
| Education Level | <0.001 | |||
| High (≥High School) | 1,523 (9.7%) | 1,362 (11%) | 161 (5.0%) | |
| Low (≤Middle School) | 14,254 (90%) | 11,163 (89%) | 3,091 (95%) | |
| BMI (kg/m2,mean±SD) | 23.9 ± 4.0 | 24.0 ± 3.9 | 23.9 ± 4.1 | 0.2 |
| BMI Category (WHO) | 0.020 | |||
| Normal | 9,233 (59%) | 7,336 (59%) | 1,897 (58%) | |
| Underweight | 887 (5.6%) | 672 (5.4%) | 215 (6.6%) | |
| Overweight | 4,791 (30%) | 3,841 (31%) | 950 (29%) | |
| Obese | 866 (5.5%) | 676 (5.4%) | 190 (5.8%) | |
| Smoking Status | <0.001 | |||
| Never | 9,063 (57%) | 6,934 (55%) | 2,129 (65%) | |
| Current/Former | 6,714 (43%) | 5,591 (45%) | 1,123 (35%) | |
| Alcohol Status | <0.001 | |||
| Never | 8,529 (54%) | 6,619 (53%) | 1,910 (59%) | |
| Current/Former | 7,248 (46%) | 5,906 (47%) | 1,342 (41%) | |
| Hypertension | <0.001 | |||
| No Hypertension | 10,750 (68%) | 8,700 (69%) | 2,050 (63%) | |
| Hypertension | 5,027 (32%) | 3,825 (31%) | 1,202 (37%) | |
| Diabetes | <0.001 | |||
| No Diabetes | 14,300 (91%) | 11,430 (91%) | 2,870 (88%) | |
| Diabetes | 1,477 (9.4%) | 1,095 (8.7%) | 382 (12%) |
Analysis of the Nonlinear Relationship Between BMI and LBP
To further investigate the nonlinear relationship between BMI and low back pain, we employed RCS analysis. The test for overall association yielded a P value < 0.001, indicating a significant association between BMI and low back pain. The test for nonlinearity produced a P value of 0.003, suggesting a significant nonlinear trend in this relationship. According to the RCS-fitted curve, BMI and the risk of low back pain exhibit a complex nonlinear relationship: when BMI is below 23.9 kg/m2, the risk is relatively high; within the 23.9–33.8 kg/m2 range, the risk is lower; and when BMI exceeds 33.8 kg/m2, the risk shows a gradual upward trend. These findings indicate that BMI may serve as an important predictor of low back pain risk, particularly with a markedly increased risk at lower BMI levels (Figure 8A).
Figure 8.
Relationship between BMI and LBP. (A) RCS plot showing a significant U-shaped nonlinear observational association between BMI and LBP prevalence (P overall <0.001, P nonlinearity=0.003). Red line = OR; blue shade = 95% CI; dashed line = OR=1. (B) MR results of the association between BMI and LBP. Point estimates and 95% CIs of OR from five MR approaches; bolded P-values indicate statistical significance.
Results of Sensitivity Analysis
CCA included 12,789 participants (2,710 with LBP), compared with 15,777 participants (3,252 with LBP) in the MI primary analysis. Key estimates were nearly identical between the two methods: LBP prevalence (21.2% in CCA vs 20.6% in MI), mean BMI (23.88 vs 23.94 kg/m2), and obesity proportion (5.4% vs 5.5%). The categorical BMI–LBP association did not reach statistical significance in CCA (P = 0.2) compared with MI (P = 0.020). This change reflects the expected reduction in statistical power due to a 19% smaller sample size and a slightly smaller absolute difference in obesity prevalence between LBP and non-LBP groups (0% in CCA vs 0.3% in MI). The direction and magnitude of all point estimates remained unchanged, confirming that missing data did not introduce substantial bias (Table 2).
Table 2.
Basic Characteristics of the Population in CHARLS (CCA - Sensitivity Analysis)
| Total Sample (n=12,789) | No LBP (N = 10,079) | LBP (N = 2710) | P-Value | |
|---|---|---|---|---|
| Age (years,mean±SD) | 61.4 ± 9.5 | 61.2 ± 9.5 | 61.8 ± 9.5 | <0.001 |
| Sex | <0.001 | |||
| Male | 5,947 (47%) | 5,051 (50%) | 896 (33%) | |
| Female | 6,842 (53%) | 5,028 (50%) | 1,814 (67%) | |
| Education Level | <0.001 | |||
| High (≥High School) | 1,407 (11%) | 1,260 (13%) | 147 (5.4%) | |
| Low (≤Middle School) | 11,382 (89%) | 8,819 (87%) | 2,563 (95%) | |
| BMI (kg/m2,mean±SD) | 23.9 ± 4.0 | 23.9 ± 4.0 | 23.8 ± 4.1 | 0.2 |
| BMI Category (WHO) | 0.2 | |||
| Normal | 7,525 (59%) | 5,941 (59%) | 1,584 (58%) | |
| Underweight | 759 (5.9%) | 574 (5.7%) | 185 (6.8%) | |
| Overweight | 3,813 (30%) | 3,019 (30%) | 794 (29%) | |
| Obese | 692 (5.4%) | 545 (5.4%) | 147 (5.4%) | |
| Smoking Status | <0.001 | |||
| Never | 5,561 (43%) | 4,591 (46%) | 970 (36%) | |
| Current/Former | 7,228 (57%) | 5,488 (54%) | 1,740 (64%) | |
| Alcohol Status | <0.001 | |||
| Never | 5,821 (46%) | 4,705 (47%) | 1,116 (41%) | |
| Current/Former | 6,968 (54%) | 5,374 (53%) | 1,594 (59%) | |
| Hypertension | <0.001 | |||
| No Hypertension | 8,554 (67%) | 6,896 (68%) | 1,658 (61%) | |
| Hypertension | 4,235 (33%) | 3,183 (32%) | 1,052 (39%) | |
| Diabetes | <0.001 | |||
| No Diabetes | 11,555 (90%) | 9,176 (91%) | 2,379 (88%) | |
| Diabetes | 1,234 (9.6%) | 903 (9.0%) | 331 (12%) |
Results of MR Analysis
Two-sample Mendelian randomization analysis indicated a significant causal relationship between high BMI and LBP, with an odds ratio (OR) of 1.396, a 95% CI of 1.232–1.581, and p < 0.001. Furthermore, the direction of effect of other methods was consistent with the IVW analysis (ORs all > 1), which provides strong support for the reliability of the study’s conclusions. The heterogeneity test results showed a p-value greater than 0.05, suggesting no significant heterogeneity among the instrumental variables. This non-significant result indicates that the selected 184 SNPs, as instrumental variables, exhibited a consistent effect in estimating the causal relationship between high BMI and LBP risk. The p-value for the MR-Egger intercept test was 0.252, further supporting the robustness of the results. A p-value greater than 0.05 indicates no horizontal pleiotropy, meaning that the selected SNPs do not affect LBP through other pathways independent of high BMI (Figure 8B).
Discussion
LBP represents a significant burden on global public health, yet the extent of this burden remains underrecognized. Previous research, focusing on specific regional analyses, lacks comprehensiveness and depth, failing to fully reveal the global impact of LBP.23 This study presents a comprehensive analysis of trends in key modifiable risk factors for LBP globally and across five SDI regions using data from the GBD, CHARLS, and GWAS databases. Specifically, we used ARIMA models to project future burden to 2041 and joinpoint regression to analyze temporal trends from 1990 to 2021. First, our study found that with advances in prevention and intervention policy strategies, the age-standardized DALYs rate attributable to smoking showed a rapid decline from 2021 to 2041, and DALYs rates related to environmental or occupational risks also improved. However, high BMI, as a major risk factor for LBP, is projected to exhibit an increasing disease burden over the next 20 years. Specifically, high BMI shows a marked increase globally, across the five SDI regions, and in multiple regions (such as high-middle SDI, middle SDI, North Africa and the Middle East, Southeast Asia, East Asia, and Oceania), with these areas experiencing the fastest percentage increases. Second, environmental or occupational risks are the second leading risk factor for LBP disease burden; over the next 20 years, the age-standardized DALYs rate is rising more rapidly in low SDI and high-middle SDI regions, while it is projected to decline in other regions. Although some areas (eg, high SDI regions) show declining age-standardized rates, the total number of DALYs cases continues to increase. Third, by age-sex trends, we found that LBP patients are mainly concentrated among middle-aged and older adults and are more prevalent in female patients. In addition, we found a significant nonlinear relationship between BMI and LBP risk among adults aged 45 years and older, indicating that both low and elevated BMI may increase the risk of LBP. Of note, the higher LBP risk associated with low BMI in the CHARLS analysis may reflect reduced muscle mass, insufficient spinal support, malnutrition, or residual confounding. Reverse causation should also be considered: chronic LBP may restrict physical activity and lead to unintended weight loss, which could contribute to the observed association at low BMI levels. Crucially, these three analyses serve different inferential purposes: GBD provides population-level burden forecasts (prediction), CHARLS identifies a cross-sectional association with potential reverse causation (association), and MR supports causality only for genetic liability to higher BMI, not for the U-shape. These findings provide novel insights for the management and prevention of LBP.
In recent years, with the improvement of healthcare systems and advancements in social development, many countries worldwide have continued to increase resource investment in chronic disease prevention, and a series of targeted measures have begun to show results in the intervention and management of LBP. For example, the United Kingdom has significantly increased the coverage of rehabilitation services for LBP patients, ensuring that more patients can obtain professional rehabilitation support in a timely manner; the United States has promoted chronic disease prevention programs at scale, attracting more than 3 million participants and effectively improving public awareness of LBP prevention and self-management capabilities; Germany has focused on upgrading medical hardware, enhancing device performance and expanding the network of rehabilitation facilities, substantially improving the accuracy of early diagnosis and the standardization of treatment for LBP; Australia, relying on long-term physical therapy programs, has reduced the average rehabilitation time for LBP patients by approximately 20% over the past decade, markedly improving patient prognosis.24 However, despite these measures having optimized LBP intervention systems, the overall global burden of LBP continues to rise, with the problem particularly pronounced among middle-aged and older populations. Research indicates that with increasing age, the musculoskeletal system function of middle-aged and older adults progressively declines: on the one hand, muscle strength in the lower back and legs decreases significantly, by about 2% per year on average, directly weakening spinal support; on the other hand, bone mineral density continues to decline, averaging 1–2% per year in women and 0.5–1% per year in men, further increasing the risk of spinal injury.25 At the same time, reduced muscle strength and decreased ligament elasticity lower this group’s tolerance for poor posture, making pressures from prolonged static postures or repetitive strain more readily transmit to the spine, thereby inducing or aggravating LBP symptoms. Against this backdrop, LBP has become one of the leading causes of nonfatal disability among middle-aged and older populations.
It is noteworthy that changes in modes of social production have introduced new challenges for the prevention and control of LBP. As demand for traditional manual labor declines, prolonged sitting and sustained static postures have become the predominant lifestyles in modern populations, a shift that markedly increases the risk of LBP.26 Studies have found that in low- and middle-income countries, working populations have about a 2.5-fold higher risk of chronic LBP compared with non-working populations, indicating that occupational factors remain a key area for LBP prevention and control.27 In fact, LBP is not only the second most common reason for clinical consultations but also the leading cause of work-related disability.2 Multiple studies have confirmed that improper lifting posture, frequent bending, prolonged stooping, maintaining static postures for excessive periods, and high-intensity physical labor all significantly increase the likelihood of LBP, further underscoring the importance of LBP prevention and control in public health.11
To date, lifestyle changes such as prolonged sitting, high-calorie, high-fat, and high-sugar diets, and reduced physical activity have led to a year-by-year increase in the global number of individuals with obesity; obesity, in turn, further reduces population physical activity levels, creating a vicious cycle that ultimately significantly raises the risk of obesity-related health problems.28–30 Moreover, studies have found that in some developing and low-income countries (such as China, Southeast Asia, and regions of Africa), the rate of increase in low back pain and spinal burden is particularly pronounced. The core drivers of these trends lie in socio-economic development–driven changes in labor structure, which have made sedentary lifestyles the norm and markedly reduced physical activity, thereby simultaneously increasing the incidence risk of high BMI and LBP.31,32 This study further confirms that the incidence among female patients is significantly higher than that among males. This difference is largely associated with estrogen deficiency after menopause—loss of estrogen’s protective effects directly leads to a decrease in basal metabolic rate, which in turn significantly increases body weight and elevates the prevalence of abdominal obesity. Existing research indicates that, in middle-aged and older populations, the higher obesity rate among women compared with men further exacerbates the risk of insulin resistance, increases the likelihood of developing type 2 diabetes, and ultimately results in a heavier disease burden.33 It is noteworthy that population aging is also a key risk factor that cannot be ignored. With advancing age, women experience a sustained decline in estrogen levels, accompanied by physiological changes such as bone mass loss and decreased muscle strength, making them more prone to lower back pain, intervertebral disc herniation, and fractures; these health issues, in turn, adversely affect physical activity capacity and create new health risks.34–36LBP can lead to reduced physical activity, deconditioning, and subsequent weight loss, rather than low BMI causing LBP. Second, low BMI in older adults often reflects frailty and sarcopenia which are independently associated with spinal instability and LBP. Third, chronic diseases or malnutrition may lead to both low BMI and increased LBP risk through systemic inflammation or reduced bone density. Fourth, unmeasured confounding could contribute to both low BMI and LBP.
Given the rising burden of lower back pain, countries need to develop and implement targeted social policies and public health interventions based on their own stage of development, forming systematic solutions. First, given the projected continued increase in LBP burden attributable to high BMI – particularly in middle and low SDI regions – interventions targeting healthy weight maintenance deserve priority. Concrete multi-level strategies should be rolled out across diverse populations, including population-wide fiscal policies such as sugar-sweetened beverage taxes to curb excessive caloric intake, community-based standardized physical activity programs, and standardized clinical weight management services delivered at primary care facilities for overweight or obese adults living with concurrent low back pain. Supplementary supporting measures cover workplace activity breaks, subsidized nutritional counselling, and multidisciplinary lifestyle intervention protocols combining dietary modification with low-impact spine-friendly exercise, with special implementation focus on groups with high sedentary prevalence.37,38 Second, although smoking-attributable LBP burden is declining globally, it remains substantial in some regions (eg, Montenegro and Serbia); continued tobacco control efforts are therefore justified. Third, the burden from environmental and occupational risks remains dominant in low SDI regions, suggesting that ergonomic improvements and workplace interventions should be prioritized there, rather than broad infrastructure investments.39 In addition, our age-sex analysis revealed a consistently higher LBP burden among females across all three risk factors. This finding suggests that sex-specific considerations may be relevant for future LBP prevention strategies. However, the underlying mechanisms remain unclear from our data. Therefore, we cautiously suggest that public health monitoring and future research should pay greater attention to sex differences in LBP burden, particularly in relation to high BMI and aging.
Notably, within the GBD analytical framework, environmental/occupational risks, smoking, and high BMI collectively account for only approximately 47% of the total global burden attributable to LBP. This indicates that more than half of LBP burden arises from other understudied risk factors not included in the present analysis. This observation underscores the incompleteness of our current understanding of LBP etiology. Accordingly, sustained and in-depth future research is warranted to elucidate additional contributors to LBP burden, including psychosocial stress, sleep disturbances, genetic susceptibility, and comorbid chronic conditions, so as to establish a more holistic framework for LBP prevention and clinical management.
This study has several limitations. First, the ARIMA prediction model is built upon historical data and operates under several key assumptions: that the time series is stationary (or can be made stationary through differencing), that future values are a linear function of past values and past forecast errors, and that the patterns observed from 1990 to 2021 will continue into the forecast period (2021–2041) in the absence of major disruptions. However, these assumptions may not hold in real-world settings. Sudden public health events, policy changes, or medical breakthroughs are unpredictable and could substantially alter existing trends, thereby affecting prediction accuracy. While we present 95% uncertainty intervals to reflect statistical uncertainty, these intervals do not capture the potential impact of unforeseen structural breaks. Therefore, the projected trends should be interpreted as conditional scenarios rather than definitive forecasts. Second, the data used in this study mainly come from the GBD, CHARLS, and GWAS databases. Differences in the diagnosis and data collection procedures for LBP across these sources may introduce bias. In particular, LBP status in the CHARLS dataset was based on participant self-report rather than clinical verification, which is susceptible to recall bias and social desirability bias, potentially leading to non-differential or differential misclassification of LBP. This misclassification could either attenuate or distort the estimated associations. Third, the GBD data lack information on individuals with low BMI, making the specific impact of low BMI on LBP burden unclear. Future research is recommended to explore the impact of both low and high BMI on LBP using longitudinal designs, to incorporate more comprehensive body composition indicators (eg, muscle mass, fat distribution), and to conduct cross-cultural comparisons to improve generalizability.
Conclusions
Based on the GBD database, this study assesses the impact of three modifiable risk factors on the burden of LBP over the next 20 years. The results indicate that the burden of LBP attributable to smoking and occupational environmental risks will show a year-by-year decline, whereas the burden attributable to high BMI will continue to increase. The burden among female patients is significantly higher than among male patients, and the LBP burden intensifies with advancing age. Furthermore, RCS analysis reveals a nonlinear relationship between BMI and low back pain: BMI increases the risk of LBP at both lower and higher ranges.
Acknowledgments
We thank everyone who contributed to the GBD database and the CHARLS database.
Funding Statement
There is no funding to report.
Abbreviations
ASDR, Age-standardized DALYs rate; AIC, Akaike Information Criterion; APC, Annual percentage change; AAPC, Average annual percentage change; ARIMA, Autoregressive Integrated Moving Average; BMI, Body mass index; CHARLS, China Health and Retirement Longitudinal Study; CI, Confidence interval; DALYs, Disability-adjusted life years; GBD, Global Burden of Disease; GWAS, Genome-wide association study; ICD, International Classification of Diseases; IVW, Inverse-variance weighted; LBP, Low back pain; MR, Mendelian randomization; PPS, Probability-proportional-to-size; RCS, Restricted cubic spline; SDI, Socio-Demographic Index; UI, Uncertainty Interval.
Data Sharing Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Ethics Approval and Consent to Participate
This study is a secondary analysis based on anonymized information data from the publicly available CHARLS database, obtained legally. According to Article 32, Items 1 and 2, of the “Ethical Review Measures for Life Science and Medical Research Involving Human Subjects in China” (February 18, 2023), this research is exempt from ethical approval under the national legislative guidelines. And the ethical approval of CHARLS is provided by the Institutional Review Board (IRB) of Peking University (IRB00001052-11015), which is updated annually. All participants provided informed consent before joining the survey. We confirm that all methods were conducted in accordance with relevant guidelines and regulations.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors declare that there are no conflicts of interest to report. We affirm that our research was conducted without any personal or financial relationships that could be perceived as influencing the results or interpretations presented in this manuscript.
References
- 1.Kabeer AS, Osmani HT, Patel J, Robinson P, Ahmed N. The adult with low back pain: causes, diagnosis, imaging features and management. British Journal of Hospital Medicine. 2023;84(10):1–20. doi: 10.12968/hmed.2023.0063 [DOI] [PubMed] [Google Scholar]
- 2.Edwards J, Hayden J, Asbridge M, Gregoire B, Magee K. Prevalence of low back pain in emergency settings: a systematic review and meta-analysis. BMC Musculoskeletal Disorders. 2017;18(1):143. doi: 10.1186/s12891-017-1511-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Agha RA, Rashid R, Kerwan A, et al. Transparency In The reporting of Artificial INtelligence – the TITAN guideline. Premier J Sci. 2025;10:100082. [Google Scholar]
- 4.Chou R. Low Back Pain. Annals of Internal Medicine. 2021;174(8):ITC113–ITC28. doi: 10.7326/AITC202108170 [DOI] [PubMed] [Google Scholar]
- 5.Hooten WM, Cohen SP. Evaluation and Treatment of Low Back Pain. Mayo Clinic Proceedings. 2015;90(12):1699–1718. doi: 10.1016/j.mayocp.2015.10.009 [DOI] [PubMed] [Google Scholar]
- 6.Itz CJ, Geurts JW, van Kleef M, Nelemans P. Clinical course of non‐specific low back pain: a systematic review of prospective cohort studies set in primary care. European Journal of Pain. 2012;17(1):5–15. doi: 10.1002/j.1532-2149.2012.00170.x [DOI] [PubMed] [Google Scholar]
- 7.Panagopoulos JHJ, Steffens D, Hancock MJ. Do MRI Findings Change Over a Period of Up to 1 Year in Patients With Low Back Pain and/or Sciatica?: a Systematic Review. Spine. 2017;42(7):504–512. doi: 10.1097/BRS.0000000000001790 [DOI] [PubMed] [Google Scholar]
- 8.Jn K. Lumbar disc disorders and low-back pain: socioeconomic factors and consequences. J Bone Joint Surg Am. 2006;88(Suppl 2):21–24. doi: 10.2106/JBJS.E.01273 [DOI] [PubMed] [Google Scholar]
- 9.Hong J, Reed C, Novick D, Happich M. Costs Associated With Treatment of Chronic Low Back Pain. Spine. 2013;38(1):75–82. doi: 10.1097/BRS.0b013e318276450f [DOI] [PubMed] [Google Scholar]
- 10.Schofield DJ, Shrestha RN, Percival R, Passey ME, Callander EJ, Kelly SJ. The personal and national costs of early retirement because of spinal disorders: impacts on income, taxes, and government support payments. The Spine Journal. 2012;12(12):1111–1118. doi: 10.1016/j.spinee.2012.09.036 [DOI] [PubMed] [Google Scholar]
- 11.Ferreira ML, de Luca K, Haile LM, et al. Global, regional, and national burden of low back pain, 1990–2020, its attributable risk factors, and projections to 2050: a systematic analysis of the Global Burden of Disease Study 2021. The Lancet Rheumatology. 2023;5(6):e316–e29. doi: 10.1016/S2665-9913(23)00098-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhang J, Wang B, Zou C, Wang T, Yang L, Zhou Y. Low back pain trends attributable to high body mass index over the period 1990–2021 and projections up to 2036. Frontiers in Nutrition. 2025;11:1521567. doi: 10.3389/fnut.2024.1521567 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wu M, Wu P, Lu H, Han L, Liu X. Global burden of occupational ergonomic factor-induced low back pain, 1990~2021: data analysis and projections of the global burden of disease. Frontiers in Public Health. 2025;13:1573828. doi: 10.3389/fpubh.2025.1573828 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ferrari AJ, Santomauro DF, Aali A, et al. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021. The Lancet. 2024;403(10440):2133–2161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Brauer M, Roth GA, Aravkin AY, et al. Global burden and strength of evidence for 88 risk factors in 204 countries and 811 subnational locations, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021. The Lancet. 2024;403(10440):2162–2203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort Profile: the China Health and Retirement Longitudinal Study (CHARLS). International Journal of Epidemiology. 2012;43(1):61–68. doi: 10.1093/ije/dys203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wei G, Lin F, Cao C, Hu H, Han Y. Non-linear dose-response relationship between body mass index and stroke risk in middle-aged and elderly Chinese men: a nationwide Longitudinal Cohort Study from CHARLS. Frontiers in Endocrinology. 2023;14:1203896. doi: 10.3389/fendo.2023.1203896 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Nazari Kangavari HSA, Hashemi Nazari SS. Suicide Mortality Trends in Four Provinces of Iran with the Highest Mortality, from 2006-2016. J Res Health Sci. 2017;17(2):e00382. [PubMed] [Google Scholar]
- 19.Zhao X, Li C, Ding G, et al. The Burden of Alzheimer’s Disease Mortality in the United States, 1999-2018. Journal of Alzheimer’s Disease. 2021;82(2):803–813. doi: 10.3233/JAD-210225 [DOI] [PubMed] [Google Scholar]
- 20.Safiri S, Carson-Chahhoud K, Noori M, et al. Burden of chronic obstructive pulmonary disease and its attributable risk factors in 204 countries and territories, 1990-2019: results from the Global Burden of Disease Study 2019. BMJ. 2022;378:e069679. doi: 10.1136/bmj-2021-069679 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kim HJ, Feuer EJ, Midthune DN. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med. 2000;19(3):335–351. doi: 10.1002/(SICI)1097-0258(20000215)19:3<335::AID-SIM336>3.0.CO;2-Z [DOI] [PubMed] [Google Scholar]
- 22.Z G. Getting to Know Pain Effect Sizes-Guidelines for Effect Size and Sample Size in Global Pain Research. Arch Phys Med Rehabil. 2026;107(4):726–733. doi: 10.1016/j.apmr.2026.01.006 [DOI] [PubMed] [Google Scholar]
- 23.Zhang C, Qin L, Yin F, Chen Q, Zhang S. Global, regional, and national burden and trends of Low back pain in middle-aged adults: analysis of GBD 1990–2021 with projections to 2050. BMC Musculoskeletal Disorders. 2024;25(1):886. doi: 10.1186/s12891-024-08002-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Cieza A, Causey K, Kamenov K, Hanson SW, Chatterji S, Vos T. Global estimates of the need for rehabilitation based on the Global Burden of Disease study 2019: a systematic analysis for the Global Burden of Disease Study 2019. The Lancet. 2020;396(10267):2006–2017. doi: 10.1016/S0140-6736(20)32340-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lim J-Y, Frontera WR. Skeletal muscle aging and sarcopenia: perspectives from mechanical studies of single permeabilized muscle fibers. Journal of Biomechanics. 2023;152:111559. doi: 10.1016/j.jbiomech.2023.111559 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bezzina A, Austin E, Nguyen H, James C. Workplace Psychosocial Factors and Their Association With Musculoskeletal Disorders: a Systematic Review of Longitudinal Studies. Workplace Health & Safety. 2023;71(12):578–588. doi: 10.1177/21650799231193578 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Jackson T, Thomas S, Stabile V, Shotwell M, Han X, McQueen K. A Systematic Review and Meta-Analysis of the Global Burden of Chronic Pain Without Clear Etiology in Low- and Middle-Income Countries: trends in Heterogeneous Data and a Proposal for New Assessment Methods. Anesthesia & Analgesia. 2016;123(3):739–748. doi: 10.1213/ANE.0000000000001389 [DOI] [PubMed] [Google Scholar]
- 28.Buchbinder R, van Tulder M, Öberg B, et al. Low back pain: a call for action. The Lancet. 2018;391(10137):2384–2388. doi: 10.1016/S0140-6736(18)30488-4 [DOI] [PubMed] [Google Scholar]
- 29.Lentz TA, Coronado RA, Master H. Delivering Value Through Equitable Care for Low Back Pain: a Renewed Call to Action. Journal of Orthopaedic & Sports Physical Therapy. 2022;52(7):414–418. doi: 10.2519/jospt.2022.10815 [DOI] [PubMed] [Google Scholar]
- 30.Wong CKW, Mak RYW, Kwok TSY, 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. The Journal of Pain. 2022;23(4):509–534. doi: 10.1016/j.jpain.2021.07.012 [DOI] [PubMed] [Google Scholar]
- 31.Yang W-L, Jiang W-C, Peng Y-H, Zhang X-J, Zhou R. Low back pain in China: disease burden and bibliometric analysis. World Journal of Orthopedics. 2024;15(12):1200–1207. doi: 10.5312/wjo.v15.i12.1200 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Sharma S, McAuley JH. Low Back Pain in Low- and Middle-Income Countries, Part 1: the Problem. Journal of Orthopaedic & Sports Physical Therapy. 2022;52(5):233–235. doi: 10.2519/jospt.2022.11145 [DOI] [PubMed] [Google Scholar]
- 33.Heuch I, Heuch I, Hagen K, Storheim K, Zwart J-A. Does the risk of chronic low back pain depend on age at menarche or menopause? A population-based cross-sectional and cohort study: the Trøndelag Health Study. BMJ Open. 2022;12(2):e055118. doi: 10.1136/bmjopen-2021-055118 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Narayan S, Pratap R, Raj G, et al. Prevalence of Osteoporosis and Sarcopenia in Middle-Aged Subjects with Low Back Pain. Indian Journal of Radiology and Imaging. 2024;35(01):002–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Arita S, Ishimoto Y, Hashizume H, et al. Age-related prevalence of radiographic lumbar spondylolisthesis and its associations with low back pain, walking speed, and muscle index: findings from the second survey of the ROAD study. European Spine Journal. 2025;34(4):1359–1365. doi: 10.1007/s00586-025-08751-x [DOI] [PubMed] [Google Scholar]
- 36.Takeuchi Y, Takahashi S, Ohyama S, et al. Relationship between body mass index and spinal pathology in community-dwelling older adults. European Spine Journal. 2022;32(2):428–435. doi: 10.1007/s00586-022-07495-2 [DOI] [PubMed] [Google Scholar]
- 37.Mudd E, Davidson SRE, Kamper SJ, et al. Healthy Lifestyle Care vs Guideline-Based Care for Low Back Pain: a Randomized Clinical Trial. JAMA Netw Open. 2025;8(1):e2453807. doi: 10.1001/jamanetworkopen.2024.53807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Briggs AD, Mytton OT, Kehlbacher A, Tiffin R, Rayner M, Scarborough P. Overall and income specific effect on prevalence of overweight and obesity of 20% sugar sweetened drink tax in UK: econometric and comparative risk assessment modelling study. BMJ. 2013;347:f6189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.O’Hagan ET, Cashin AG, Traeger AC, McAuley JH. Person-centred education and advice for people with low back pain: making the best of what we know. Brazilian Journal of Physical Therapy. 2023;27(1):100478. doi: 10.1016/j.bjpt.2022.100478 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.