Abstract
Background
Body pain affecting over 30% of China’s aging population, imposes significant socioeconomic burdens, yet environmental determinants remain understudied. This national study investigates associations between multidimensional living environments and pain among middle-aged and older adults.
Methods
Using 2011–2020 data from the China Health and Retirement Longitudinal Study (CHARLS), eligible participants were included in cross-sectional and longitudinal analyses. Living environmental quality was assessed via five indicators: building types, household temperatures, water sources, energy sources, and outdoor PM2.5 exposure, categorized into favorable, moderate, and unfavorable. Pain outcomes included upper limb, lower limb, trunk, head and neck, and multisite pain. Logistic and Cox regression models were utilized to examine associations between living environment quality and body pain.
Results
Cross-sectionally, unfavorable environments were linked to higher prevalence of single-site (e.g., lower limb OR: 2.05, 95% CI: 1.78–2.37) and multisite pain (OR: 2.02, 95% CI: 1.77–2.32) versus suitable environments, with significant dose-response relationships (all P-values for trend < 0.001). In longitudinal analyses, unfavorable environments increased 9-year incident pain risks: upper limb (HR: 1.30, 95% CI: 1.20–1.41), lower limb (HR: 1.39, 95% CI: 1.28–1.50), trunk (HR: 1.18, 95% CI: 1.09–1.28), head and neck (HR: 1.20, 95% CI: 1.10–1.31), and multisite pain (HR: 1.31, 95% CI: 1.21–1.42), with consistent dose-response patterns (all P-values for trend < 0.001). Kaplan-Meier curves demonstrated a significant difference in pain incidence across different living environment groups within the entire cohort (all P-values for log-rank test < 0.001).
Conclusions
Cumulative environmental exposures independently predict pain incidence among middle-aged and older adults in China. Targeted upgrades to the living environment could reduce pain burdens, informing healthy aging policies in China and similar settings.
Supplementary Information
The online version contains supplementary material available at 10.1186/s13690-025-01732-y.
Keywords: CHARLS, Cohort, Living environmental quality, Pain
| Text box 1. Contributions to the literature |
|---|
| 1. There is a dose-response relationship between living environment quality and body pain. |
| 2. Residents living in unfavorable environment could be key targets for body pain prevention. |
| 3. Findings support public health policies aimed at improving living environment quality, which may help mitigate the pain burden. |
Introduction
The world is undergoing unprecedented population aging, with projections indicating that the global population aged ≥ 60 years will surpass 2.1 billion by 2050, predominantly residing in developing countries [1]. In China, accelerated demographic transitions have positioned middle-aged and older adults as a demographic confronting distinct health challenges, particularly marked by an elevated prevalence of chronic pain [2, 3]. Chronic pain, widely recognized as a leading contributor to disability and reduced quality of life in aging populations, has emerged as a critical global public health priority [4]. Recent epidemiological studies estimate that over 30% of middle-aged and older Chinese adults experience persistent musculoskeletal or generalized pain [2, 5]. The socioeconomic burden attributable to pain is substantial, encompassing both direct healthcare expenditures and indirect productivity losses [6]. While individual-level determinants, such as advanced age [7], physical inactivity [8], chronic disease comorbidities [9], and socioeconomic disadvantage [7] have been well-documented, growing evidence implicates environmental exposures—particularly modifiable living conditions—as pivotal yet understudied contributors to pain pathogenesis [10].
Low- and middle-income countries (LMICs), including China, confront distinct environmental challenges characterized by accelerated urbanization [11], pervasive ambient air pollution [12], and deficient housing infrastructure [13]. Epidemiological investigations have established that chronic exposure to air pollutants—primarily nitrogen dioxide (NO2) and ozone (O3)—exacerbates joint pain sensitivity in patients with rheumatoid arthritis [14], suggesting systemic pathways for pain exacerbation. Indoor environmental quality parameters, particularly non-optimal temperature and humidity conditions, demonstrate significant associations with increased risks of low back pain [15]. Additional evidence indicates that prolonged consumption of untreated water [16] and residence in substandard housing infrastructures (e.g., poorly constructed prefabricated units) [17] may potentiate susceptibility to musculoskeletal disorders and chronic pain. Despite these findings, few studies have integrated these multifactorial environmental exposures into a unified analytical framework, especially within aging populations in developing countries.
China’s vast population size and distinct urbanization patterns provide an advantageous context for investigating these relationships. Utilizing nationally representative data from the China Health and Retirement Longitudinal Study (CHARLS), this study examines associations between multifaceted living environment indicators—including building types, household temperature, water sources, energy sources, and outdoor PM2.5 exposure—and self-reported body pain among middle-aged and older adults. By adopting a holistic environmental exposure framework, our findings aim to inform targeted public health strategies to alleviate preventable pain burden through environmental modifications, ultimately advancing health equity in aging populations.
Methods
Study design and population
This study integrated cross-sectional and longitudinal designs using data from the CHARLS [18]. As a nationally database, CHARLS collects health status and socioeconomic characteristics data among populations aged 45 years or above through standardized protocols, thereby providing robust interdisciplinary support for aging-related research. Researchers employed multistage probability-proportional-to-size (PPS) sampling strategy, which covered 150 counties or urban districts across 28 provinces. Initiated in 2011, the project has released four waves of follow-up data in 2013, 2015, 2018, and 2020. The study protocol was approved by the Biomedical Ethics Committee of Peking University (Approval: IRB00001052-11015), and written informed consent was obtained from all participants before data collection.
In this study, baseline data were used to assess environmental exposures and pain status (2011). Pain outcomes—including upper limb, lower limb, trunk, head-and-neck, and multisite pain—were tracked through four waves of follow-up data (2013–2020). For cross-sectional analyses, we excluded 421 participants aged below 45 years, 329 with missing environmental data, and 62 lacking pain-related information, resulting in a sample of 16,893 participants. In longitudinal analyses, individuals with pain at baseline and those missing follow-up data were further excluded. The final participant selection process is illustrated in Fig. 1.
Fig. 1.
Flow diagram for participants included in the study
Assessment of living environmental quality
Living environmental quality was assessed through five core indicators: building types, household temperatures, water sources, energy sources for cooking and heating, and outdoor PM2.5 exposure [19–21]. The scoring method and algorithm have been detail tabulated in Table S1, each environmental factor was dichotomously coded as follows: (1) Building types: multi-story structures = 0, single-story structures = 1; (2) Household temperatures: categorized based on interviewer-subjective evaluation as suitable = 0 or very hot, hot, cold, and very cold = 1; (3) Water sources: tap water = 0, non-tap water = 1; (4) Energy sources for cooking and heating: Households using clean fuels (e.g., solar, natural gas, liquefied petroleum gas, electricity) for both activities = 0; those using solid fuels (e.g., coal, crop residue, wood) for both activities = 2; and those with mixed usage (one clean, one solid) = 1; and (5) Outdoor PM2.5 exposure: annual mean concentrations were classified per standards of the Ministry of Ecology and Environment of China as < 35.0 µg/m³ = 0 or ≥ 35.0 µg/m³ = 1. PM2.5 concentration data were derived from the China High Air Pollutants (CHAP) database [22], which integrates multi-source observations including ground monitoring networks, satellite remote sensing products, atmospheric reanalysis, and model simulations to establish a high-resolution (1-km grid) exposure assessment system. Given individual exposure data privacy regulations, access was limited to city-level PM2.5 data, which served as a proxy for individual-level data [23].
The composite living environment score (range: 0–6) was calculated by summing the dichotomous codes, with higher scores indicating a greater cumulative burden of adverse environmental factors. Consistent with the established epidemiological framework [19, 20], the environment was classified into three levels: favorable (0–1 points), moderate (2–3 points), and unfavorable (4–6 points).
Assessment of pain
This study operationalized body pain as primary outcome measures. Pain characteristics were assessed using standardized self-report questions: “Are you often troubled with any pains? On what part of your body do you feel pain? Please list all body areas where you are currently experiencing pain.” Based on anatomical distribution, 15 pain sites were categorized into four primary regions [24]: (1) upper limb (shoulder, arm, wrist, and fingers), (2) lower limb (leg, knees, ankle, and toes), (3) trunk (chest, stomach, back, waist, and buttocks), and (4) head and neck. Multisite pain was defined as concurrent pain occurring in two or more distinct anatomical regions [25].
Covariates
Based on prior knowledge [5, 26, 27], the following covariates were considered: sociodemographic characteristics included age, sex, residence (rural, urban), marital status (married, others), and educational status (illiteracy, primary school and below, and middle school and above; lifestyle habits included smoking status (never smokers, ever smokers) and drinking status (never drinkers, ever drinkers); chronic disease included hypertension, dyslipidemia, diabetes, cancer, chronic lung diseases, liver disease, heart attack, stroke, kidney disease, stomachache, psychiatric disease, memory-related disease, arthritis, and asthma. Chronic conditions were ascertained through self-reported physician-diagnosed information, assessed by the question: “Have you been diagnosed with any of the following conditions by a physician?“.
Statistical analyses
Data were described as mean ± standard deviation or frequency and percentage. Baseline participant characteristics were compared across groups defined by living environment quality, employing Chi-square tests for categorical data and one-way ANOVA for normally distributed continuous data as appropriate. Missing covariates (Table S2) were addressed through multiple imputations via chained equations [28], generating 50 imputed datasets. The most plausible dataset was selected for the final analysis. In cross-sectional analyses, logistic regression models were employed to estimate associations between living environment quality and pain, with results expressed as odds ratios (ORs) and 95% confidence intervals (CIs). While for longitudinal analyses, Cox proportional hazards models with follow-up time as the timescale were constructed, reporting hazard ratios (HRs) and 95% CIs. Follow-up time was measured as the interval between the first interview and either the pain event recording or the most recent interview. The proportional hazards assumption was assessed using Schoenfeld residuals. For covariates violating this assumption, stratification approach was implemented to address non-proportionality. Four models were constructed: Model 1 was an unadjusted model; Model 2 was adjusted for age, sex, residence, marital status, and education status; Model 3 was further adjusted for smoking and drinking status; and Model 4 was additionally adjusted for chronic disease conditions.
Kaplan-Meier curves were utilized to estimate cumulative incidence of pain stratified by living environment quality, between-group differences assessed via log-rank tests. Subgroup and interaction analyses were conducted to explore potential effect modifications by sociodemographic characteristics. To evaluate the robustness of findings, three sensitivity analyses were performed: First, analyses were repeated using complete-case datasets to mitigate potential bias from missing variables; Second, new-onset pain cases identified during the first follow-up wave were excluded to minimize confounding effects on causal associations; Third, neighborhood- and city-level measurements may violate the model’s independence assumption. We evaluated potential clustering effects at these levels by computing intraclass correlation coefficients (ICC).
All statistical analyses were conducted using SPSS version 25.0 and Stata version 17.0. A two-sided P-value < 0.05 was considered statistically significant throughout the study.
Results
Characteristics of study population
Table 1 presents the baseline characteristics of 16,893 participants according to living environmental quality. The study population had a mean age of 59.4 ± 9.9 years, with 8,248 participants (48.8%) being male. Compared to participants residing in suitable environments, those in unfavorable environments were significantly older and had higher proportions of rural residence, unmarried individuals, lower educational attainment, and current smokers. Furthermore, participants exposed to worse environments exhibited elevated prevalence of comorbidities including hypertension, dyslipidemia, diabetes, chronic lung diseases, heart attack, kidney disease, stomachache, psychiatric disease, memory-related disease, arthritis, and asthma (all P-values < 0.05).
Table 1.
Baseline characteristics of participants according to living environmental quality
| Characteristics | Total | Living environmental quality | P-value | ||
|---|---|---|---|---|---|
| Suitable | Moderate | Unfavorable | |||
| N (%) | 16,893 | 3194 (18.9) | 5536 (23.8) | 8163 (48.3) | |
| Age, years, mean (SD) | 59.4 ± 9.9 | 58.6 ± 9.9 | 58.8 ± 9.8 | 60.1 ± 9.9 | < 0.001 |
| Sex, n (%) | 0.581 | ||||
| Male | 8248 (48.8) | 1534 (48.0) | 2705 (48.9) | 4009 (49.1) | |
| Female | 8645 (51.2) | 1660 (52.0) | 2831 (51.1) | 4154 (50.9) | |
| Residence, n (%) | < 0.001 | ||||
| Rural | 13,128 (77.7) | 1314 (41.1) | 4341 (78.4) | 7473 (91.5) | |
| Urban | 3765 (22.3) | 1880 (58.9) | 1195 (21.6) | 690 (8.5) | |
| Marital status, n (%) | < 0.001 | ||||
| Married | 13,532 (80.1) | 2634 (82.5) | 4450 (80.4) | 6448 (79.0) | |
| Others | 3361 (19.9) | 560 (17.5) | 1086 (19.6) | 1715 (21.0) | |
| Educational status, n (%) | < 0.001 | ||||
| Illiteracy | 4638 (27.5) | 426 (13.3) | 1422 (25.7) | 2790 (34.2) | |
| Primary school and below | 6639 (39.3) | 973 (30.5) | 2277 (41.1) | 3389 (41.5) | |
| Middle school and above | 5616 (33.2) | 1795 (56.2) | 1837 (33.2) | 1984 (24.3) | |
| Drinking status, n (%) | 0.905 | ||||
| Ever drinkers | 5620 (33.3) | 1053 (33.0) | 1851 (33.4) | 2716 (33.3) | |
| Never drinkers | 11,273 (66.7) | 2141 (67.0) | 3685 (66.6) | 5447 (66.7) | |
| Smoking status, n (%) | < 0.001 | ||||
| Ever smokers | 6791 (40.2) | 1130 (35.4) | 2164 (39.1) | 3497 (42.8) | |
| Never smokers | 10,102 (59.8) | 2064 (64.6) | 3372 (60.9) | 4666 (57.2) | |
| Chronic disease, n (%) | |||||
| Hypertension | 4172 (24.7) | 856 (26.8) | 1325 (23.9) | 1991 (24.4) | 0.008 |
| Dyslipidemia | 1554 (9.2) | 438 (13.7) | 482 (8.7) | 634 (7.8) | < 0.001 |
| Diabetes | 978 (5.8) | 247 (7.7) | 339 (6.1) | 392 (4.8) | < 0.001 |
| Cancer | 168 (1.0) | 33 (1.0) | 59 (1.1) | 76 (0.9) | 0.716 |
| Chronic lung diseases | 1750 (10.4) | 251 (7.9) | 457 (8.3) | 1042 (12.8) | < 0.001 |
| Liver disease | 665 (3.9) | 109 (3.4) | 206 (3.7) | 350 (4.3) | 0.059 |
| Heart attack | 2033 (12.0) | 418 (13.1) | 546 (9.9) | 1069 (13.1) | < 0.001 |
| Stroke | 402 (2.4) | 89 (2.8) | 116 (2.1) | 197 (2.4) | 0.120 |
| Kidney disease | 1103 (6.5) | 181 (5.7) | 338 (6.1) | 584 (7.2) | 0.005 |
| Stomachache | 3764 (22.3) | 562 (17.6) | 1196 (21.6) | 2006 (24.6) | < 0.001 |
| Psychiatric disease | 236 (1.4) | 30 (0.9) | 57 (1.0) | 149 (1.8) | < 0.001 |
| Memory-related disease | 265 (1.6) | 54 (1.7) | 63 (1.1) | 148 (1.8) | 0.006 |
| Arthritis | 5634 (33.4) | 796 (24.9) | 1782 (32.2) | 3056 (37.4) | < 0.001 |
| Asthma | 627 (3.7) | 83 (2.6) | 179 (3.2) | 365 (4.5) | < 0.001 |
Cross-section association between living environmental quality and pain
Table S3 presents the prevalence of pain stratified by living environmental quality. Poorer living environments were associated with an increased prevalence of both single-site (including upper limb, lower limb, trunk, and head and neck) pain and multisite pain (all P-values < 0.001). Detailed associations between living environments and pain are provided in Table 2. Results from continuous variable models revealed positive correlations between cumulative exposure to unfavorable environmental factors and the prevalence of single-site and multisite pain. In categorical variable models, a dose-response relationship was observed between worsening living environments and pain prevalence. Specifically, compared to participants in suitable environments, those in moderate environments had a higher prevalence of single-site and multi-site pain, with further increases observed in unfavorable environments. Table S4 identifies specific environmental factors significantly associated with pain, included residence in single-story buildings, exposure to unfavorable household temperature, non-tap water source, and the use of solid fuels for cooking and heating (all P-values < 0.001).
Table 2.
Cross-sectional association between living environmental quality and pain
| Living environmental scores (OR, 95% CI) | P-value for trend | ||||
|---|---|---|---|---|---|
| Continuous variable | Suitable (0–1) | Moderate (2–3) | Unfavorable (4–6) | ||
| Upper limb pain | |||||
| Model 1 | 1.27 (1.24, 1.31) | Reference | 1.77 (1.54, 2.03) | 2.62 (2.30, 2.99) | < 0.001 |
| Model 2 | 1.21 (1.17, 1.25) | Reference | 1.45 (1.27, 1.70) | 2.05 (1.77, 2.36) | < 0.001 |
| Model 3 | 1.21 (1.17, 1.25) | Reference | 1.46 (1.26, 1.70) | 2.03 (1.76, 2.34) | < 0.001 |
| Model 4 | 1.16 (1.12, 1.19) | Reference | 1.40 (1.20, 1.63) | 1.76 (1.52, 2.05) | < 0.001 |
| Lower limb pain | |||||
| Model 1 | 1.32 (1.29, 1.36) | Reference | 1.95 (1.71, 2.23) | 3.06 (2.70, 3.47) | < 0.001 |
| Model 2 | 1.24 (1.21, 1.28) | Reference | 1.59 (1.38, 1.83) | 2.31 (2.02, 2.65) | < 0.001 |
| Model 3 | 1.24 (1.20, 1.28) | Reference | 1.59 (1.38, 1.83) | 2.29 (2.00, 2.63) | < 0.001 |
| Model 4 | 1.20 (1.16, 1.24) | Reference | 1.54 (1.33, 1.79) | 2.05 (1.78, 2.37) | < 0.001 |
| Trunk pain | |||||
| Model 1 | 1.28 (1.24, 1.31) | Reference | 1.70 (1.51, 1.91) | 2.58 (2.32, 2.88) | < 0.001 |
| Model 2 | 1.21 (1.18, 1.25) | Reference | 1.42 (1.26, 1.61) | 2.04 (1.81, 2.30) | < 0.001 |
| Model 3 | 1.21 (1.18, 1.24) | Reference | 1.42 (1.25, 1.60) | 2.03 (1.80, 2.28) | < 0.001 |
| Model 4 | 1.17 (1.13, 1.20) | Reference | 1.36 (1.19, 1.55) | 1.78 (1.57, 2.03) | < 0.001 |
| Head and neck pain | |||||
| Model 1 | 1.29 (1.25, 1.33) | Reference | 1.77 (1.51, 2.02) | 2.68 (2.34, 3.08) | < 0.001 |
| Model 2 | 1.23 (1.19, 1.27) | Reference | 1.45 (1.25, 1.70) | 2.12 (1.82, 2.46) | < 0.001 |
| Model 3 | 1.23 (1.19, 1.27) | Reference | 1.45 (1.24, 1.69) | 2.11 (1.82, 2.45) | < 0.001 |
| Model 4 | 1.17 (1.13, 1.22) | Reference | 1.39 (1.19, 1.64) | 1.83 (1.56, 2.13) | < 0.001 |
| Multisite pain | |||||
| Model 1 | 1.32 (1.28, 1.35) | Reference | 1.87 (1.65, 2.11) | 2.97 (2.64, 3.33) | < 0.001 |
| Model 2 | 1.25 (1.21, 1.29) | Reference | 1.53 (1.34, 1.75) | 2.28 (2.01, 2.59) | < 0.001 |
| Model 3 | 1.25 (1.21, 1.28) | Reference | 1.53 (1.34, 1.74) | 2.27 (2.00, 2.58) | < 0.001 |
| Model 4 | 1.20 (1.17, 1.24) | Reference | 1.48 (1.29, 1.70) | 2.02 (1.77, 2.32) | < 0.001 |
OR, odds ratio; CI, confidence interval
Model 1 was not conducted with any adjustment; Model 2 was adjusted for age, sex, residence, educational status, and marital status; Model 3 was further adjusted for smoking status and drinking status; Model 4 was additionally adjusted for hypertension, dyslipidemia, diabetes, cancer, chronic lung diseases, liver disease, heart attack, stroke, kidney disease, stomachache, psychiatric disease, memory-related disease, arthritis, and asthma
Longitudinal association between living environmental quality and pain
Table S5 shows that during the median follow-up of 9.0 years, 5,712 incident upper limb, 5,883 lower limb, 6,011 trunk, 4,998 head and neck, and 5,879 multisite pain cases were recorded. Compared with participants in suitable environments, those residing in moderate and unfavorable environments exhibited higher incidence rates for both single-site and multisite pain (all P-values < 0.001). Table 3 demonstrates that longitudinal and cross-sectional results were essentially consistent, indicating a graded association between living environmental quality and incident pain. When environmental scores were modeled as continuous variables, fully adjusted analyses revealed significant positive associations with risks of both single-site and multisite pain. In categorical models, versus suitable environments, exposure to moderate environments was associated with increased risks for new-onset upper limb pain, lower limb pain, and multisite pain, while unfavorable environments showed further elevated risks for all single-site and multisite pain. Specific environmental hazards were identified in Table S6, including residence in single-story buildings, exposure to unfavorable household temperatures, non-tap water sources, and solid fuel usage, whereas outdoor PM2.5 exposure demonstrated a protective effect (all P-values < 0.05).
Table 3.
Longitudinal association between living environmental quality and pain
| Living environmental scores (HR, 95% CI) | P-value for trend | ||||
|---|---|---|---|---|---|
| Continuous variable | Suitable (0–1) | Moderate (2–3) | Unfavorable (4–6) | ||
| Upper limb pain | |||||
| Model 1 | 1.11 (1.09, 1.13) | Reference | 1.22 (1.12, 1.32) | 1.53 (1.42, 1.65) | < 0.001 |
| Model 2 | 1.09 (1.07, 1.11) | Reference | 1.13 (1.03, 1.22) | 1.39 (1.28, 1.51) | < 0.001 |
| Model 3 | 1.09 (1.07, 1.11) | Reference | 1.12 (1.03, 1.22) | 1.39 (1.28, 1.51) | < 0.001 |
| Model 4 | 1.08 (1.05, 1.10) | Reference | 1.10 (1.02, 1.20) | 1.33 (1.22, 1.45) | < 0.001 |
| Lower limb pain | |||||
| Model 1 | 1.13 (1.11, 1.15) | Reference | 1.24 (1.15, 1.34) | 1.59 (1.48, 1.71) | < 0.001 |
| Model 2 | 1.11 (1.09, 1.13) | Reference | 1.19 (1.09, 1.29) | 1.49 (1.37, 1.61) | < 0.001 |
| Model 3 | 1.11 (1.09, 1.13) | Reference | 1.19 (1.10, 1.29) | 1.49 (1.38, 1.62) | < 0.001 |
| Model 4 | 1.10 (1.08, 1.12) | Reference | 1.18 (1.08, 1.28) | 1.44 (1.33, 1.56) | < 0.001 |
| Trunk pain | |||||
| Model 1 | 1.09 (1.07, 1.11) | Reference | 1.17 (1.08, 1.26) | 1.39 (1.30, 1.50) | < 0.001 |
| Model 2 | 1.06 (1.04, 1.09) | Reference | 1.08 (1.00, 1.17) | 1.25 (1.15, 1.35) | < 0.001 |
| Model 3 | 1.06 (1.04, 1.09) | Reference | 1.08 (0.99, 1.17) | 1.25 (1.15, 1.35) | < 0.001 |
| Model 4 | 1.06 (1.04, 1.08) | Reference | 1.06 (0.98, 1.15) | 1.21 (1.12, 1.31) | < 0.001 |
| Head and neck pain | |||||
| Model 1 | 1.11 (1.08, 1.13) | Reference | 1.19 (1.10, 1.30) | 1.44 (1.33, 1.56) | < 0.001 |
| Model 2 | 1.08 (1.06, 1.10) | Reference | 1.10 (1.01, 1.20) | 1.30 (1.19, 1.41) | < 0.001 |
| Model 3 | 1.08 (1.06, 1.10) | Reference | 1.09 (1.01, 1.20) | 1.30 (1.19, 1.41) | < 0.001 |
| Model 4 | 1.07 (1.04, 1.09) | Reference | 1.07 (0.97, 1.17) | 1.23 (1.13, 1.35) | < 0.001 |
| Multisite pain | |||||
| Model 1 | 1.12 (1.10, 1.14) | Reference | 1.21 (1.12, 1.30) | 1.54 (1.43, 1.65) | < 0.001 |
| Model 2 | 1.10 (1.08, 1.12) | Reference | 1.13 (1.05, 1.23) | 1.41 (1.30, 1.53) | < 0.001 |
| Model 3 | 1.10 (1.08, 1.12) | Reference | 1.13 (1.05, 1.23) | 1.40 (1.30, 1.52) | < 0.001 |
| Model 4 | 1.09 (1.07, 1.11) | Reference | 1.12 (1.03, 1.22) | 1.36 (1.26, 1.48) | < 0.001 |
HR, hazard ratio; CI, confidence interval
Model 1 was not conducted with any adjustment; Model 2 was adjusted for age, sex, residence, educational status, and marital status; Model 3 was further adjusted for smoking status and drinking status; Model 4 was additionally adjusted for hypertension, dyslipidemia, diabetes, cancer, chronic lung diseases, liver disease, heart attack, stroke, kidney disease, stomachache, psychiatric disease, memory-related disease, arthritis, and asthma
Figure 2 shows the Kaplan–Meier curves for cumulative incidence of single-site pain across environmental exposure strata. The suitable environment group exhibited the lowest cumulative incidence, whereas the unfavorable environment group showed the highest cumulative incidence, with a statistically significant difference observed between groups (all P-values for log-rank test < 0.001).
Fig. 2.
Kaplan–Meier curves for cumulative pain risk by living environmental quality
Subgroup and sensitivity analyses
Tables S7-S11 present subgroup analyses stratified by age, sex, residence, educational status, and marital status. The associations between living environmental quality and body pain remained consistent across all pre-specified subgroups, with no significant effect modification observed (all P-values for interaction > 0.05). Sensitivity analyses using complete-case datasets showed concordant results with multiple imputation approaches (Table S12). After excluding incident cases identified in the second wave (Table S13), the results remained largely unchanged. As shown in Table S14, clustering effects at both city and neighborhood levels were minimal, suggesting limited violation of the independence assumption. These support the robustness of our primary findings.
Discussions
This large-scale national study provides compelling evidence that suboptimal living environment quality is independently associated with elevated risks of incident pain across multiple anatomical regions in Chinese middle-aged and older adults. Both cross-sectional and 9-year follow-up analyses revealed a consistent dose-response relationship, with individuals exposed to unfavorable environments exhibiting significantly higher risks of developing pain compared to those in suitable living conditions. Notably, specific environmental hazards—encompassing solid fuel use, non-tap water dependence, single-story building, and uncomfortable household temperatures—all showed independent associations with pain outcomes. These findings persisted across diverse sociodemographic subgroups and retained robust after rigorous adjustment for chronic comorbidities and lifestyle factors.
Our study aligns with emerging evidence linking air pollution to pain amplification. Ziadé et al. [14] documented exacerbated joint pain extent in rheumatoid arthritis patients exposed to nitrogen dioxide and ozone, while Vodonos et al. [29] reported increased headache-related emergency departments visits during high-pollution periods. Mediation analyses suggested that indoor air pollution from solid fuel use exacerbated chronic pain risks in middle-aged and older adults [30]. Mechanistically, air pollutants may induce nociceptive sensitization through oxidative stress pathways and pro-inflammatory cytokine release (e.g., IL-6, TNF-α), which amplify peripheral nociceptor activation and promote chronic pain [31–33]. Notably, the protective effect of PM2.5 observed in supplementary analyses warrants caution—city-level exposure metrics may mask individual heterogeneity, and residual confounding (e.g., socioeconomic status and healthcare service facilities accessibility) could distort associations [12, 34].
Substandard building structures confer elevated risks of chronic pain, with single-story residents exhibiting 12%-16% higher limb pain incidence. This finding aligns with disaster epidemiology evidence showing prefabricated housing occupants face increased susceptibility to new-onset chronic musculoskeletal disorders [17]. The housing-chronic pain association involves multifactorial mechanisms, potentially representing a bio-psycho-social pain syndrome through integrated pathway [35]. In contrast, multi-story buildings typically feature superior ventilation and thermal regulation systems [20]. Substandard housing often lacks centralized heating, amplifying cold stress exposure—a known trigger of musculoskeletal loading and ischemic pain [36]. Matsugaki et al. [15] identified cold indoor environments as an independent risk factor for teleworkers’ low back pain, while Lewis et al. [37] reported 1.6-fold greater neck-shoulder pain incidence among workers with occupational cold exposure. In addition, structural vulnerabilities in aging houses (e.g., poor ventilation, dampness) likely contributes to inflammatory arthritis flares, thereby promote body pain [21].
The elevated risk of multi-regional pain associated with non-tap water dependence extends findings from occupational cohorts exposed to untreated water. Workers relying on untreated water sources demonstrated a 51% higher risk of incident musculoskeletal disorders compared to those consuming clean drinking water [16]. Contaminants in unregulated water sources (e.g., heavy metals, arsenic, fluorides, and pathogens) may induce chronic pain via gut microbiome disruption or direct tissue toxicity [38, 39]. For instance, fluoride accumulation in bones and joints has been implicated in skeletal fluorosis, a painful condition particularly endemic to regions with groundwater contamination [40]. Furthermore, water-fetching burdens in areas lacking tap water infrastructure (common in rural China) likely compounds mechanical stress on spine and extremities, potentially constituting a major contributor to disability-associated musculoskeletal disorders [41].
Our integrated exposure framework highlights the multifactorial nature of environmentally driven pain, revealing how substandard housing, indoor air pollutants, thermal discomfort, and untreated water collectively exacerbate pain risks, particularly in resource-constrained regions. The environmental risk scoring system enables the precise identification of vulnerable populations, guiding infrastructure targeted upgrades aligned with China’s Rural Revitalization Strategy [42] and Dual-Carbon Policy [43], including prioritized tap water provision in western provinces and clean heating subsidies for northeastern households. Architectural interventions should further incorporate optimized thermal regulation and enhanced humidity control systems into housing design. Such interventions could transform living environments into proactive health promoters rather than passive disease determinants, potentially alleviating pain burden in aging populations while decelerating progression of sarcopenia [42], arthritis [21] and cognitive decline [20].
The strengths of this study include the nationally representative CHARLS cohort, longitudinal design, and comprehensive environmental exposure assessment. Nevertheless, several limitations warrant consideration. First, the living environment assessments relied on self-reported measures, potentially introducing recall bias. Second, although we focused on baseline exposure-pain associations, environmental quality is subject to temporal variability. Future studies should therefore incorporate repeated exposure measurements to capture dynamic changes in environmental quality and assess the potential impact of such temporal alterations on pain outcomes. Third, ambient temperature classification relied on subjective interviewer assessments during household visits. While capturing contextual exposure relevant to perceived comfort, this approach does not quantify objective temperature levels (e.g., seasonal changes may influence subjective judgments.) or independently evaluate differences in the magnitude of body pain associated with heat versus cold exposure. Future studies would benefit from incorporating standardized, continuous temperature measurements. Fourth, environmental systems operate holistically, but our analyses examined specific components, offering a limited perspective. Furthermore, urban-level PM2.5 estimates might not sufficiently capture individual exposure variability and cannot represent all airborne pollutants.
Conclusion
This study demonstrates that cumulative environmental burdens constitute a modifiable determinant of pain incidence among middle-aged and older adults in China. The identified dose-response relationships underscore the necessity for integrated environmental health policies targeting solid fuel usage, substandard housing infrastructure, and untreated water sources. By prioritizing these interventions, policymakers could substantially mitigate preventable pain burden and advance healthy aging agendas in LMICs.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
The authors sincerely thank the China Health and Retirement Longitudinal Study team for providing data and training on using the datasets. The authors also extend their gratitude to all volunteers and staff involved in this research.
Abbreviations
- CHAP
China High Air Pollutants
- CHARLS
China Health and Retirement Longitudinal Study
- CI
Confidence interval
- HR
Hazard ratio
- LMICs
Low- and middle-income countries
- NO2
Nitrogen dioxide
- O3
Ozone
- OR
Odds ratio
- PPS
Probability-proportional-to-size
Author contributions
S.H. was involved in conceptualization, formal analysis, investigation, visualization, writing—original draft, and writing—review and editing. X.Z. was involved in formal analysis, writing—original draft, and writing—review and editing. R.L. was involved in conceptualization, finding acquisition, supervision, and writing—review and editing. All authors have read and approved the final version of the manuscript.
Funding
This study was supported by the Xi’an Science and Technology Plan Project (22YXYJ0125). The recipients of this funding is the corresponding author Rui Liu.
Data availability
The data that support the findings of this study are available from CHARLS project site, subject to registration and application process. CHARLS datasets are available for download at the CHARLS home website (http://charls.pku.edu.cn).
Declarations
Ethics approval and consent to participate
The entire study process adhered to the Declaration of Helsinki, and the study results were reported following the STROBE guidelines. The protocol for the CHARLS cohort was authorized by the Ethics Review Committee of Peking University (IRB00001052–11015), and all participants provided written informed consent at the time of participation.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Shuanglong Hou and Xin Zhao contributed equally to this work.
References
- 1.Rudnicka E, Napierała P, Podfigurna A, Męczekalski B, Smolarczyk R, Grymowicz M. The world health organization (WHO) approach to healthy ageing. Maturitas. 2020;139:6–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Qiu Y, Li H, Yang Z, Liu Q, Wang K, Li R, et al. The prevalence and economic burden of pain on middle-aged and elderly Chinese people: results from the China health and retirement longitudinal study. BMC Health Serv Res. 2020;20(1):600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wang Y, Chen B, Liu X, Zeng H, Chen B, Wang Z, et al. Temporal trends in the burden of musculoskeletal diseases in China from 1990 to 2021 and predictions for 2021 to 2030. Bone. 2025;191:117332. [DOI] [PubMed] [Google Scholar]
- 4.Henschke N, Kamper SJ, Maher CG. The epidemiology and economic consequences of pain. Mayo Clin Proc. 2015;90(1):139–47. [DOI] [PubMed] [Google Scholar]
- 5.Yang L, Peng W. Prevalence and factors associated with body pain: results of a nationally representative survey of 9,586 Chinese adults aged 60 and over. Front Public Health. 2021;9:634123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hadjiat Y, Serrie A, Treves R, Chomier B, Geranton L, Billon S. Pain associated with health and economic burden in france: results from recent National health and wellness survey data. Clinicoecon Outcomes Res. 2018;10:53–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mills SEE, Nicolson KP, Smith BH. Chronic pain: a review of its epidemiology and associated factors in population-based studies. Br J Anaesth. 2019;123(2):e273–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dzakpasu FQS, Carver A, Brakenridge CJ, Cicuttini F, Urquhart DM, Owen N, et al. Musculoskeletal pain and sedentary behaviour in occupational and non-occupational settings: a systematic review with meta-analysis. Int J Behav Nutr Phys Act. 2021;18(1):159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Alsubaie SF, Alkathiry AA, Aljuaid MI, Alnasser MA. The relationship between chronic diseases and the intensity and duration of low back pain. Eur J Phys Rehabil Med. 2024;60(1):55–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sant’Anna MB, Kimura LF, Vieira WF, Zambelli VO, Novaes LS, Hösch NG, et al. Environmental factors and their impact on chronic pain development and maintenance. Phys Life Rev. 2024;48:176–97. [DOI] [PubMed] [Google Scholar]
- 11.Xie X, Que J, Sun L, Sun T, Yang F. Association between urbanization levels and frailty among middle-aged and older adults in china: evidence from the CHARLS. BMC Med. 2025;23(1):171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wang Y, Wang Y, Xu H, Zhao Y, Marshall JD. Ambient air pollution and socioeconomic status in China. Environ Health Perspect. 2022;130(6):67001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Nie P, Li Y, Ding L, Sousa-Poza A. Housing poverty and healthy aging in china: evidence from the China health and retirement longitudinal study. Int J Environ Res Public Health. 2021;18(18). [DOI] [PMC free article] [PubMed]
- 14.Ziadé N, Bouzamel M, Mrad-Nakhlé M, Abi Karam G, Hmamouchi I, Abouqal R, et al. Prospective correlational time-series analysis of the influence of weather and air pollution on joint pain in chronic rheumatic diseases. Clin Rheumatol. 2021;40(10):3929–40. [DOI] [PubMed] [Google Scholar]
- 15.Matsugaki R, Muramatsu K, Tateishi S, Nagata T, Tsuji M, Hino A, et al. Association between telecommuting environment and low back pain among Japanese telecommuting workers: A Cross-Sectional study. J Occup Environ Med. 2021;63(12):e944–8. [DOI] [PubMed] [Google Scholar]
- 16.Saha A, Mukherjee AK, Ravichandran B. Musculoskeletal problems and fluoride exposure: A cross-sectional study among metal smelting workers. Toxicol Ind Health. 2016;32(9):1581–8. [DOI] [PubMed] [Google Scholar]
- 17.Sogi Y, Yabe Y, Hagiwara Y, Sekiguchi T, Sugawara Y, Tsuchiya M, et al. Association between continued residence in temporary prefabricated housing and musculoskeletal pain in survivors of the great East Japan earthquake: a longitudinal study. BMJ Open. 2019;9(10):e030761. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol. 2014;43(1):61–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Yang Y, Cao L, Xia Y, Li J. The effect of living environmental factors on cardiovascular diseases in Chinese adults: results from a cross-sectional and longitudinal study. Eur J Prev Cardiol. 2023;30(11):1063–73. [DOI] [PubMed] [Google Scholar]
- 20.Luo H, Hu H, Zheng Z, Sun C, Yu K. The impact of living environmental factors on cognitive function and mild cognitive impairment: evidence from the Chinese elderly population. BMC Public Health. 2024;24(1):2814. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Liu R, Zhou Y, Liu Y, Guo R, Gao L. Association between living environmental quality and risk of arthritis in middle-aged and older adults: a National study in China. Front Public Health. 2023;11:1181625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wei J, Li Z, Lyapustin A, Sun L, Peng Y, Xue W, et al. Reconstructing 1-km-resolution high-quality PM2.5 data records from 2000 to 2018 in china: Spatiotemporal variations and policy implications. Remote Sens Environ. 2021;252:112136. [Google Scholar]
- 23.Wen Y, Zhu G, Cao K, Liang J, Lu X, Wang T. The association between PM(2.5) and frailty: evidence from 122 cities in China and 7 countries in Europe. BMC Public Health. 2024;24(1):3612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Huang H, Ni L, Zhang L, Zhou J, Peng B. Longitudinal association between frailty and pain in three prospective cohorts of older population. J Nutr Health Aging. 2025;29(6):100537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Chen J, Yan L, Chu J, Wang X, Xu Z. Pain characteristics and progression to sarcopenia in Chinese Middle-Aged and older adults: A 4-Year longitudinal study. J Gerontol Biol Sci Med Sci. 2024;79(5). [DOI] [PMC free article] [PubMed]
- 26.Ai Z, Tang C, Peng P, Wen X, Tang S. Prevalence and influencing factors of chronic pain in middle-aged and older adults in china: results of a nationally representative survey. Front Public Health. 2023;11:1110216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Luo J, Ma Y, Zhan HW, Jia WH, Zhang JR, Xie SY, et al. Associations between adverse childhood experiences and pain in middle-aged and older adults: findings from the China health and retirement longitudinal study. BMC Public Health. 2024;24(1):1760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377–99. [DOI] [PubMed] [Google Scholar]
- 29.Vodonos A, Novack V, Zlotnik Y, Ifergane G. Ambient air pollution, weather and daily emergency department visits for headache. Cephalalgia. 2015;35(12):1085–91. [DOI] [PubMed] [Google Scholar]
- 30.Jia Z, Gao Y, Zhao L, Han S. Effects of pain and depression on the relationship between household solid fuel use and disability among middle-aged and older adults. Sci Rep. 2022;12(1):21270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Fandiño-Del-Rio M, Kephart JL, Williams KN, Malpartida G, Boyd Barr D, Steenland K, et al. Household air pollution and blood markers of inflammation: A cross-sectional analysis. Indoor Air. 2021;31(5):1509–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Molot J, Sears M, Marshall LM, Bray RI. Neurological susceptibility to environmental exposures: pathophysiological mechanisms in neurodegeneration and multiple chemical sensitivity. Rev Environ Health. 2022;37(4):509–30. [DOI] [PubMed] [Google Scholar]
- 33.Lane KJ, Levy JI, Patton AP, Durant JL, Zamore W, Brugge D. Relationship between traffic-related air pollution and inflammation biomarkers using structural equation modeling. Sci Total Environ. 2023;870:161874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hanley O, Miner J, Rockswold E, Biros M. The relationship between chronic illness, chronic pain, and socioeconomic factors in the ED. Am J Emerg Med. 2011;29(3):286–92. [DOI] [PubMed] [Google Scholar]
- 35.Yabuki S, Ouchi K, Kikuchi S, Konno S. Pain, quality of life and activity in aged evacuees living in temporary housing after the great East Japan earthquake of 11 March 2011: a cross-sectional study in Minamisoma city, Fukushima Prefecture. BMC Musculoskelet Disord. 2015;16:246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Oksa J, Ducharme MB, Rintamäki H. Combined effect of repetitive work and cold on muscle function and fatigue. J Appl Physiol (1985). 2002;92(1):354–61. [DOI] [PubMed] [Google Scholar]
- 37.Lewis C, Stjernbrandt A, Wahlström J. The association between cold exposure and musculoskeletal disorders: a prospective population-based study. Int Arch Occup Environ Health. 2023;96(4):565–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Namkaew M, Wiwatanadate P. Association of fluoride in water for consumption and chronic pain of body parts in residents of San Kamphaeng district, Chiang mai, Thailand. Trop Med Int Health. 2012;17(9):1171–6. [DOI] [PubMed] [Google Scholar]
- 39.Kile ML, Rodrigues EG, Mazumdar M, Dobson CB, Diao N, Golam M, et al. A prospective cohort study of the association between drinking water arsenic exposure and self-reported maternal health symptoms during pregnancy in Bangladesh. Environ Health. 2014;13(1):29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Srivastava S, Flora SJS. Fluoride in drinking water and skeletal fluorosis: a review of the global impact. Curr Environ Health Rep. 2020;7(2):140–6. [DOI] [PubMed] [Google Scholar]
- 41.Geere JA, Bartram J, Bates L, Danquah L, Evans B, Fisher MB, et al. Carrying water May be a major contributor to disability from musculoskeletal disorders in low income countries: a cross-sectional survey in South africa, Ghana and Vietnam. J Glob Health. 2018;8(1):010406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Lin J, Gong K, Chen C. Towards integrated sustainability for china’s rural revitalization: an analysis of income inequality and public health. Front Public Health. 2023;11:1328821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Li R, Luo Y, Li Y, Zhu X, Zhang J, Wang Z, et al. Synergistic reduction in air pollutants and health benefits under china’s Dual-Carbon policy. Environ Sci Technol. 2024;58(22):9467–70. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from CHARLS project site, subject to registration and application process. CHARLS datasets are available for download at the CHARLS home website (http://charls.pku.edu.cn).