The association between greenness exposure and arthritis in middle-aged and older Chinese adults, mediated by climate factors

Background

Arthritis is a common health issue among middle-aged and older adults, significantly impacting their quality of life. While previous studies have explored various risk factors for arthritis, the relationship between green space exposure and arthritis risk remains underexplored. This study aims to investigate the correlation between green space exposure, as measured by the Normalized Difference Vegetation Index (NDVI), and arthritis risk among middle-aged and older adults in China using a cross-sectional approach.

Methods

Data for the present study were extracted from the 2015 wave of the China Health and Retirement Longitudinal Study (CHARLS), focusing specifically on middle-aged and older adults aged 45 years and above. Greenness exposure was quantified using the NDVI. Generalized linear models were used to assess the association between NDVI and arthritis. Climatic variables (relative humidity, precipitation) and metabolic equivalents were evaluated as correlates and potential mediators of this relationship.

Results

The study included a total of 7,985 participants, of whom 3,519 had arthritis and 4,466 did not. In the fully adjusted model, NDVI showed a positive correlation with arthritis. Specifically, the odds ratio (OR) of arthritis for each interquartile range (IQR) increase in NDVI was 1.14 (95% CI: 1.02–1.27). Additionally, annual precipitation, annual relative humidity, and metabolic equivalents all showed positive associations with arthritis prevalence. Further mediation analysis indicated that annual precipitation significantly mediated the relationship between NDVI and arthritis, with a proportion mediated of 5.31%.

Conclusion

Higher NDVI was associated with a higher prevalence of arthritis, and annual precipitation partly explained this association. The findings suggest that environmental factors, including greenery and climate, may be considered in future strategies aimed at understanding and addressing joint disease burdens.

Ethics approval and consent to participate

The study was approved by the Institutional Review Board of Peking University (Code: IRB00001052-11015) and conducted in accordance with the Declaration of Helsinki, with written informed consent obtained from all participants.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26813-7.

Key Messages

• Based on CHARLS 2015 data, a per IQR increase in NDVI was associated with a 14% higher prevalence of arthritis.

• Mediation analysis indicated that annual precipitation partially accounted for the observed association between NDVI and arthritis prevalence.

• Regions with higher levels of green space may consider integrating environmental factors, such as greenery and climate, into joint disease prevention and management strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26813-7.

Background

Arthritis is a general term encompassing a variety of arthritic conditions, characterized by inflammation of the joints that can manifest as either acute or chronic presentations. It is predominantly marked by joint pain and stiffness, which can substantially compromise patients’ physical function and quality of life [1, 2]. Meanwhile, the prevalence of arthritis has reached a higher level worldwide. The incidence of arthritis increases markedly with age, particularly after 45 years. Data from 2010 to 2012 revealed that 22.7% of the US population had doctor-diagnosed arthritis, with projections suggesting a 49% increase in prevalence by 2040 [3]. Similarly, in China, the overall prevalence of arthritis among middle-aged and senior individuals grew from 31.4% to 44.7% between 2011 and 2018 [4]. Furthermore, research has shown that a substantial rise in medical expenditures has failed to enhance the health-related quality of life for individuals suffering from arthritis [5]. Considering the medical context, the importance of identifying relevant risk factors to prevent the onset of arthritis cannot be overstated.

The etiology and pathogenesis of arthritis are highly complex and involve multiple factors. To date, genetics and environmental factors are among the key influences on the onset and progression of the condition [6–8]. In recent years, most relevant studies have concentrated on environmental pollution. Research has indicated that smoking and elevated levels of air pollution are significant risk factors for the development of arthritis [8]. Their primary pathogenic mechanisms may involve negatively altering the immune response. Furthermore, studies have shown that weather and seasonal changes can mitigate disease severity [9]. In clinical practice, many patients with rheumatoid arthritis (RA) report that their joint symptoms fluctuate with changes in climatic factors, such as temperature, humidity, and atmospheric pressure. However, strong scientific evidence to support this assumption is lacking [10].

As global urbanization continues to rise, and the adverse effects of long-term exposure to urban environmental risks have garnered significant attention, prompting extensive research into the potential health outcomes associated with exposure to green spaces. The most common marker for greenness exposure is NDVI [11]. Currently, numerous aspects of greenness are beneficial. For instance, exposure to greenness has beneficial effects in most studies regarding respiratory mortality, lung cancer incidence, respiratory hospitalisations and pulmonary function [12]. Similarly, greenness has been shown to improve air quality and enhance green spaces, thereby promoting visual health in aging populations [13]. While some aspects may be harmful, studies have also emphasized greenness may have different health effects in different population subgroups [12]. Emerging evidence suggests a synergistic interaction between ambient greenness and climatic variables [14], coupled with consistent observations that greater residential greenness promotes higher levels of physical activity [15]. Concomitantly, both meteorological parameters and the intensity of physical activity have been causally implicated in the pathogenesis and progression of arthritis [16, 17]. Synthesising these converging lines of evidence, this study conducted a cross-sectional analysis focusing on the relationship between green environment exposure and arthritis among middle-aged and elderly individuals in China. Furthermore, the study examined the mediating roles of climate factors (relative humidity, precipitation) and physical activity intensity (metabolic equivalents) in the association between green environment exposure and arthritis, thereby providing robust support for etiological research into the disease.

Methods

Study population

Data for this study were derived from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative survey of adults aged 45 and older. The study was conducted with the informed consent of all participants and received approval from the Institutional Review Board of Peking University (Code: IRB00001052-11015). The baseline survey utilized a multi-stage, stratified, probability-proportional-to-size sampling design, with further details on recruitment and sampling methods provided in accordance with the study protocol [18]. A total of 17,708 participants from 10,257 households across 28 provinces were recruited between 2011 and 2012, with subsequent follow-ups conducted in 2013, 2015, and 2018.

For this cross-sectional study utilizing data from the 2015 wave of the China Health and Retirement Longitudinal Study (CHARLS), we aimed to examine the association between green space exposure and arthritis and to assess the mediating roles of climate factors and physical activity intensity. The analysis was restricted to participants aged 45 years or older. We further excluded individuals with: (1) missing data on arthritis diagnosis; (2) missing residential geolocation, which prevented assignment of green space exposure; (3) missing data on any key covariate; or (4) cognitive or physical impairment recorded by the interviewer that could compromise the reliability of survey responses. After applying these criteria, a total of 7,985 participants were included in the final analysis (Fig. 1).

Fig. 1.

Flowchart of participants selection

Greenness exposure

To assess greenness exposure, we used NDVI, a widely used satellite-derived metric for quantifying vegetation density and photosynthetic activity. NDVI is calculated as (NIR − R) / (NIR + R), where NIR and R represent near-infrared and red reflectance, respectively. The index yields values ranging from − 1 to 1, with higher positive values indicating denser and healthier vegetation [19]. We obtained NDVI data from the Moderate Resolution Imaging Spectroradiometer (MODIS) aboard NASA’s Terra satellite (MOD13Q1 product). The annual mean NDVI for 2015 was computed for each Chinese province and assigned to participants according to their residential location.

Climate factors

For this study, we assembled 2015 climate data encompassing relative humidity and precipitation. Relative humidity grids were obtained from the Geospatial Remote-sensing and Ecological Network (https://www.gisrs.cn), whereas precipitation indices were retrieved from the National Tibetan Plateau Scientific Data Center (https://data.tpdc.ac.cn). Consistent with our NDVI processing protocol, annual means were computed for each variable to attenuate seasonal fluctuations.

Physical activity assessment

The CHARLS database records respondents’ weekly physical activities, including intensity, duration, and frequency. Activities are categorized as light, moderate, and vigorous, with corresponding MET values assigned using the IPAQ scoring system [20]. MET minutes/week are calculated as follows:

• Light activity: 3.3 × (minutes × days).

• Moderate activity: 4.0 × (minutes × days).

• Vigorous activity: 8.0 × (minutes × days).

Arthritis assessment

Participants were defined as having physician - diagnosed arthritis if they answered “yes” to the question “Have you ever been diagnosed with arthritis or rheumatism by a doctor?” [21]. The assessment did not differentiate between subtypes such as osteoarthritis or rheumatoid arthritis, but aimed to capture the overall burden of joint-related chronic conditions in the population.

Covariates

Baseline data on covariates were collected by trained interviewers using a structured questionnaire. Variable distributions are detailed in Supplementary Table S1. Covariates included:

1. Sociodemographic characteristics: Gender, age, residence, marital status, and education.

2. Lifestyle factors: Smoking, alcohol use, cooking fuel, and household expenditure.

3. Anthropometric measurements: Recorded measurements included body mass index (BMI), waist circumference, and systolic and diastolic blood pressure (SBP and DBP).

4. Laboratory data: Triglycerides, high-density lipoprotein (HDL), fasting glucose, and C-reactive protein (CRP).

Statistical analysis

We used descriptive statistics to summarize baseline characteristics (mean ± SD or median with IQR for continuous variables, counts with proportions for categorical variables) and assessed group differences with t-tests or chi-square/Fisher’s exact tests. Generalized Linear Models explored the NDVI–arthritis link: Model 1 was unadjusted, Model 2 included socioeconomic and lifestyle factors, and Model 3 added anthropometric and biological indicators. Results are shown as odds ratios (OR) with confidence intervals (CI). Mediation analysis using the R package “mediation” evaluated climate factors and physical activity aspotential mediators, with bootstrapping (1,000 replications) to estimate mediation effects. Stratified analyses examined NDVI effects across subgroups defined by gender, residence, marital status, education, smoking, drinking, and cooking fuel use, using fully adjusted models to account for confounders.

All statistical analyses were conducted using R version 4.4.3, which is available at https://www.r-project.org/. Two-sided p-values less than 0.05 were considered statistically significant.

Results

The baseline characteristics of study participants

Table 1 presents the summary statistics of the study participants. Continuous variables are shown as mean with standard deviation (SD) or median with IQR, and categorical variables are presented as counts with percentages. Our final analysis included 7,985 participants, with a mean age of 61.34 ± 9.75 years. Among these, 4117 (51.56%) were female, 6453 (80.88%) were currently married or living with a partner, and 4964 (62.17%) resided in rural areas. Approximately two - thirds (67.25%) of the participants had an educational level of elementary school or below. Anthropometric measurements and laboratory test results for all participants are summarized in Supplementary Table S2.

Table 1.

Baseline Characteristics of the Study Participant

Characteristics	Total (n = 7985)	Non-Arthritis (n = 4466)	Arthritis (n = 3519)	p
Age (years), mean (SD)	61.34 (9.75)	60.55 (9.83)	62.33 (9.54)	< 0.001
BMI (kg/m2), mean (SD)	24.83 (33.14)	24.39 (15.17)	25.37 (46.54)	0.24
Gender, n(%)				< 0.001
Female	4117 (51.56)	2104 (47.11)	2013 (57.20)
Male	3868 (48.44)	2362 (52.89)	1506 (42.80)
Residence, n(%)				< 0.001
Rural	4964 (62.17)	2618 (58.62)	2346 (66.67)
Urban	3021 (37.83)	1848 (41.38)	1173 (33.33)
Marital_status, n(%)				< 0.001
Married and living with a spouse	6453 (80.88)	3703 (82.95)	2750 (78.26)
Married but living without a spouse	433 ( 5.43)	237 ( 5.31)	196 ( 5.58)
Single, divorced, and windowed	1092 (13.69)	524 (11.74)	568 (16.16)
Education_Status, n(%)				< 0.001
Elementary school or below	5370 (67.25)	2743 (61.42)	2627 (74.65)
Middle school or above	2615 (32.75)	1723 (38.58)	892 (25.35)
NDVI, mean (SD)	0.00 (0.76)	-0.04 (0.77)	0.05 (0.75)	< 0.001
Annual Relative Humidity, mean (SD)	-0.04 (0.58)	-0.08 (0.58)	0.00 (0.59)	< 0.001
Annual Precipitation(mm), mean (SD)	0.10 (0.57)	0.08 (0.58)	0.12 (0.54)	0.009
Metabolic equivalent(METs), mean (SD)	0.19 (0.70)	0.16 (0.68)	0.24 (0.73)	< 0.001

SD standard deviation, BMI body mass index 

The total percentage may not equal to 100 due to rounding

Associations between NDVI and arthritis in generalized linear models

Generalized linear models demonstrated a positive association between NDVI and arthritis. Per IQR increment in NDVI corresponded to an OR of 1.16 (95%CI: 1.10–1.23) unadjusted, 1.15 (1.05–1.26) after adjustment for sociodemographic and lifestyle factors, and 1.14 (1.02–1.27) with further control for anthropometric and biological variables.

We also calculated the associations between climate factors (relative humidity, precipitation), metabolic equivalents, and arthritis under three models. The results showed that these factors were all significantly associated with arthritis (Fig. 2). After adjusting for all covariates, per IQR increase in annual relative humidity, precipitation, and metabolic equivalents was associated with higher odds of arthritis (OR = 1.45, 95%CI 1.25–1.68; OR = 1.25, 95%CI 1.08–1.45 and OR = 1.33, 95%CI:1.17–1.50; all P < 0.01). Full results of the generalized linear models are provided in Table 2.

Fig. 2.

OR (with 95% CIs) for arthritis associated with per IQR increase in exposure to annual relative humidity, Annual Precipitation and Metabolic equivalent.Abbreviations: OR, odds ratio; CI, confidence interval; IQR, interquartile range.Model 1: unadjusted for any covariates. Model 2: adjusted for gender, age, BMI, residence, marital status, education level, smoking status, drinking status, and cooking fuel use. Model 3: adjusted for all variables in Model 2 plus waist circumference, systolic and diastolic blood pressure, triglycerides, high-density lipoprotein (HDL), fasting blood glucose, and C-reactive protein (CRP)

Table 2.

Generalized linear regression for the arthritis at follow-up

	Model1		Model2		Model3
	OR(95%CI)	P	OR(95%CI)	P	OR(95%CI)	P
NDVI	1.16 (1.10, 1.23)	< 0.001	1.15 (1.05, 1.26)	< 0.05	1.14 (1.02, 1.27)	< 0.05
Annual precipitation	1.11 (1.03, 1.20)	< 0.01	1.20 (1.06, 1.35)	< 0.01	1.25 (1.08, 1.45)	< 0.01
Annual Relative Humidity	1.26 (1.16, 1.35)	< 0.001	1.37 (1.22, 1.54)	< 0.001	1.45 (1.25, 1.68)	< 0.001
Metabolic equivalent	1.17 (1.10, 1.25)	< 0.001	1.24 (1.13, 1.40)	< 0.001	1.33 (1.17, 1.50)	< 0.001

OR (with 95% CIs) for Arthritis per IQR Increase in Exposure to Environmental and Metabolic Factors Across Adjustment Models. Model 1: unadjusted; Model 2: adjusted for socio-demographic and behavioral factors; Model 3: Model 2 plus cardiometabolic and inflammatory factors

Mediation effects

As shown in Table 3, after adjusting for covariates, annual precipitation was confirmed as a significant mediator, accounting for 5.31% of the association between NDVI and arthritis [Average Causal Mediation Effect ((ACME) = 1.89e-03, 95% CI: 4.70e-04, 1.00e-02, P < 0.01]. In contrast, annual relative humidity and metabolic equivalents did not exhibit statistically significant mediation effects. The mediation pathways are illustrated in Fig. 3.

Table 3.

Mediation effects between NDVI and arthritis, Annual relative humidity, annual precipitation and metabolic equivalent as mediators

Mediator	ADE (95%CI)	ACME (95%CI)	Total effect (95%CI)	Proportion mediated
Annual precipitation	3.37e-02 (1.14e-02, 6.00e-02) **	1.89e-03 (4.70e-04, 1.00e-02) **	3.56e-02 (1.30e-02, 6.00e-02) **	5.31e-02 **
Annual relative humidity	3.71e-02 (1.54e-02, 6.00e-02) **	-1.53e-03 (-3.83e-03, 0.00)	3.56e-02 (1.44e-02, 6.00e-02) **	-4.30e-02
Metabolic equivalent	3.48e-02 (1.51e-02, 6.00e-02) ***	8.80e-04 (-6.20e-04, 1.00e-02)	3.57e-02 (1.55e-02, 6.00e-02) ***	2.46e-02

ADE   Average Direct Effect ACME   Average Causal Mediation Effect 

All models adjusted for gender, age, BMI, residence, marital status, education level, smoking status, drinking status, and cooking fuel use 

Values are presented in scientific notation with three significant figures

** P- value < 0.01

*** P- value < 0.001

Fig. 3.

Path diagram of mediation effects between NDVI and arthritis, Annual relative humidity, annual precipitation and metabolic equivalent as mediators. 95% CI in the parentheses are shown. Model control for gender, age, BMI,residence, marital status, education level, smoking status, and drinking status, cooking fuel use. ** P- value < 0.01, *** P- value < 0.001

Subgroup analysis results

Subgroup analyses (Supplementary Table S5) indicated that, across most strata, NDVI, relative humidity, precipitation, and metabolic equivalents were positively associated with arthritis risk. Notably, the association between vegetation increase and arthritis risk was stronger among individuals residing in rural areas, those with single, divorced, or widowed marital status, those with lower education levels, those living in the Eastern region, and those using solid fuel for cooking. Interestingly, the risk was also relatively higher among non-smokers and non-drinkers.

Discussion

This cross-sectional study utilized a large-scale, nationally representative sample of over 7,000 middle-aged and older adults in China to investigate the association between green space exposure, proxied by NDVI, and arthritis. Our analysis revealed a positive association between higher NDVI and arthritis prevalence. Fully adjusted models also indicated that annual relative humidity, precipitation, and physical activity intensity were independently associated with arthritis. Notably, mediation analysis suggested annual precipitation may partially explain the observed association between NDVI and arthritis.

In recent years, the impact of greenness exposure on health has garnered increasing attention, with numerous studies confirming its benefits. However, our finding of a positive association between NDVI and arthritis contrasts with some prior research. For instance, one epidemiological study of 30,684 participants in China reported that more green space was associated with a lower risk of rheumatoid arthritis (RA) [22]. This discrepancy may be explained by differences in outcome definition (our study used a broad arthritis definition vs. RA-specific), study populations (middle-aged/older adults vs. adults aged ≥ 18), and geographical coverage (national vs. a specific regional cohort). Another study using CHARLS data (pooled waves 2011–2018) found an inverse association between NDVI and arthritis [23]. The differences from our results may stem from methodological variations, including the aggregation of longitudinal data, categorization of NDVI, and a primary focus on PM2.5 interactions. Our work, by examining mediating pathways via climate factors and physical activity, underscores the complexity of this relationship and highlights the need for refined analyses that consider non-linear effects and specific mechanisms.

The overall body of evidence supports the notion that exposure to green spaces can exert beneficial effects on health, such as reducing the risk of cardiovascular diseases, mental health disorders, adverse birth outcomes, and mortality [24, 25]. However, the characteristics of green spaces and their interactions with local contexts can lead to heterogeneous health outcomes. For example, studies in China have linked higher NDVI to an increased risk of COPD [26] and have identified green spaces as a factor in the transmission dynamics of certain infectious diseases [27, 28], potentially mediated through related climatic changes.

Environmental factors, particularly climate, are implicated in arthritis etiology and symptomatology [10]. Cold and damp conditions are climatic and environmental factors associated with increased risk [29, 30]. Low temperature, high atmospheric pressure, and high humidity are significantly correlated with pain in patients with RA [31, 32]. A retrospective longitudinal study [33] utilized multivariate linear regression to explore the links between weather factors and rheumatoid arthritis (RA) symptoms. The findings showed that winter humidity and summer rainfall significantly correlated with the number of tender joints in RA patients. Notably, slight fluctuations in atmospheric pressure (AP) related to weather changes can profoundly impact human health. The influence can be attributed to both direct mechanical effects and the modulation of oxygen partial pressure [34]. These changes may directly influence the onset and progression of arthritis, as demonstrated by studies showing that variations in barometric pressure are independently associated with the severity of knee pain in osteoarthritis patients [35].

The mechanisms linking greenness exposure to arthritis remain poorly understood and are likely multifactorial, involving complex interactions between environmental, immunological, and potentially microbiological factors [9, 36–38]. Our study contributes tothis nascent field by identifying a novel statistical association and a potential mediating pathway.

Our mediation analysis, which examined humidity, precipitation, and physical activity as potential intermediaries, aligns with existing evidence on climate-arthritis links. While both humidity and precipitation initially showed significant mediation, only precipitation remained a statistically significant mediator after covariate adjustment, highlighting its distinct role in the association between NDVI and arthritis. Although residential greenness is often associated with higher physical activity levels [39–41], the mediating effect of physical activity intensity (MET) was not significant in adjusted analyses. This indicates that pathways other than physical activity, such as climatic and potentially other environmental or biological mechanisms, may be more central to explaining the observed relationship.

Our findings offer insights for practice and policy. Clinically, the observed association suggests that environmental context, particularly residential greenness coupled with high precipitation, could be considered when assessing arthritis risk in middle-aged and older adults, potentially informing targeted screening or patient education in susceptible subgroups. For public health and urban planning, this highlights the importance of integrating local climate conditions with greening strategies—for example, ensuring good drainage and sheltered activity spaces in high-rainfall regions to promote usable green space while mitigating dampness-related exposure.

Our study has several strengths. Initially, we leveraged a sizable, nationally representative cohort to amass a broad spectrum of variables, encompassing demographic traits, lifestyle behaviors, health metrics, and both physiological and biochemical markers. The comprehensive nature of this data collection enhances the reliability and precision of our findings, thereby bolstering the validity of our study’s conclusions. Secondly, our study represents the pioneering effort to establish a positive correlation between NDVI and the prevalence of arthritis, while also providing insights into the intermediary roles played by climatic factors, including humidity and precipitation, as well as metabolic equivalents. This investigation contributes novel perspectives to the understanding of how environmental determinants may influence the development of musculoskeletal conditions. Lastly, subgroup analyses were performed to further validate the robustness of our findings. This approach allowed us to explore potential heterogeneity across different demographic and clinical subgroups, thus strengthening the generalizability of our results.

However, our study also has limitations. Firstly, the cross-sectional design of the study precludes the establishment of causality between greenness exposure and arthritis. Secondly, the use of NDVI as a proxy for greenness exposure may not fully capture the complexity of green space characteristics and human interactions with these spaces. Thirdly, the study relies on self-reported data for arthritis diagnosis, which may introduce recall bias. Finally, the generalizability of our findings is limited to middle-aged and older adults in China and may not extend to other populations or age groups.

Conclusions

In summary, this national cross-sectional study found a positive association between residential greenness exposure (NDVI) and arthritis prevalence among middle-aged and older adults in China. Part of this association was statistically explained by annual precipitation. These findings highlight a potential link between environmental factors and joint health, suggesting that greenery and climate may be relevant considerations in public health. However, the non-causal nature of this study design necessitates confirmation through longitudinal research to elucidate any underlying causal pathways and mechanisms.

Abbreviations

CHARLS

China Health and Retirement Longitudinal Study

IRB

Institutional Review Board

NDVI

Normalized Difference Vegetation Index

MODIS

Moderate Resolution Imaging Spectroradiometer

MET

Metabolic Equivalents

IPAQ

International Physical Activity Questionnaire

BMI

Body Mass Index

SBP

Systolic Blood Pressure

DBP

Diastolic Blood Pressure

HDL

High-Density Lipoprotein

CRP

C-Reactive Protein

RA

Rheumatoid Arthritis

COPD

Chronic Obstructive Pulmonary Disease

AP

Atmospheric Pressure

OR

Odds Ratio

CI

Confidence Interval

ACME

Average Causal Mediation Effect

Authors’ contributions

Statement of equal contribution: All authors have contributed equally to this study.

Funding

This work was supported by grants from the Liaoning Provincial Natural Science Fund for Distinguished Young Scholars, Science and Technology Program of Liaoning Province (Project No.: 2024-MSLH-159). The study sponsor has no role in study design, data analysis and interpretation of data, the writing of manuscript, or the decision to submit the paper for publication.

Data availability

The datasets generated and analyzed during the current study are available in the China Health and Retirement Longitudinal Study (CHARLS) repository [insert specific repository link or reference here]. Access to the data can be requested through the official CHARLS website following their data access policy. The authors declare that all data supporting the findings of this study are available upon reasonable request.

Declarations

Ethics approval and consent to participate

The data for this study were derived from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative survey of adults aged 45 and older. The study was conducted with the informed consent of all participants, who were provided with detailed information about the study objectives, procedures, potential risks and benefits, and their rights to withdraw from the study at any time without penalty. Written informed consent was obtained from each participant before the start of the study. The study received approval from the Institutional Review Board of Peking University (Code: IRB00001052-11015) and was conducted in accordance with the principles of the Declaration of Helsinki.

Consent for publication

All participants provided written informed consent for the publication of anonymized data from the study. The consent form explicitly stated that the data collected would be used for research purposes and that any published results would not include any identifying information that could link the data to individual participants. The study adhered to strict confidentiality protocols to ensure the privacy and anonymity of all participants. The Institutional Review Board of Peking University (Code: IRB00001052-11015) reviewed and approved the consent form and the procedures for data publication.

Competing interests

The authors declare no competing interests.