Abstract
Background
Knee osteoarthritis and knee pain are major causes of chronic pain, functional limitation, and health-care use, but community-level evidence on their symptomatic burden and associated factors remains limited in northwest China. We aimed to estimate the burden of knee osteoarthritis and knee pain, describe pain-related functional impact, and examine demographic, clinical, lifestyle, and occupational factors associated with these outcomes among adults in Urumqi.
Methods
We conducted a community-based cross-sectional survey in Urumqi, Xinjiang, China, from June to August 2023. The primary analysis included adults aged 40 years or older with valid anthropometric and core knee outcome data. Outcomes were self-reported physician-diagnosed knee osteoarthritis, past-year knee pain, symptomatic knee osteoarthritis, and pain-related functional impact. Modified Poisson regression with robust variance was used to estimate adjusted prevalence ratios (aPRs) and 95% CIs.
Results
Among 1925 participants, the median age was 55.0 years (IQR 48.0–63.0), and 1,118 (58.1%) were men. Physician-diagnosed knee osteoarthritis was reported by 236 participants (12.3, 95% CI 10.9–13.8), past-year knee pain by 562 (29.2%, 27.2–31.3), and symptomatic knee osteoarthritis by 212 (11.0%, 9.7–12.5). Among participants with past-year knee pain, 214 of 525 reported sleep disturbance (40.8%, 36.6–45.0), 286 of 525 reported work or activity limitation (54.5%, 50.2–58.7), and 320 of 525 had sought medical care (61.0%, 56.7–65.0). Previous knee injury was strongly associated with physician-diagnosed knee osteoarthritis (aPR 3.08, 95% CI 2.37–4.00), past-year knee pain (2.01, 1.72–2.34), and symptomatic knee osteoarthritis (2.97, 2.25–3.91). Sleep disorder, overweight, obesity, high occupational mechanical load, and first-degree family history of arthritis were also associated with knee outcomes.
Conclusion
Knee osteoarthritis and knee pain were common among middle-aged and older adults in Urumqi and were frequently accompanied by sleep disturbance, activity limitation, and health-care seeking. The observed associations with previous knee injury, excess body weight, occupational mechanical load, and sleep disorder identify groups that may warrant closer community assessment; their temporal and causal roles require confirmation in longitudinal studies.
Keywords: community-based survey, functional impact, knee osteoarthritis, knee pain, modifiable risk factors
1. Introduction
Knee osteoarthritis (KOA) is a major cause of chronic pain and disability, and its burden is increasing with population ageing and rising adiposity. The knee is the most commonly affected joint, and global analyses show sustained increases in prevalence and years lived with disability, particularly among women and older adults (1–4).
The public health burden of KOA is not captured by structural disease or diagnostic history alone. Knee pain and its effects on mobility, sleep, work, and health-care use often determine the need for care and the extent of disability (5, 6). Community-level data that jointly describe physician-diagnosed KOA, recent knee pain, and pain-related functional consequences are therefore important for local service planning.
National surveys and modelling studies describe the scale and distribution of KOA, but provide less information on how diagnostic history and symptoms coexist with functional limitation and locally relevant occupational, anthropometric, sleep, injury, and comorbidity characteristics. Such evidence remains limited in northwest China, where population ageing, work patterns, digital access, and routes to health care may differ from those in more extensively studied regions.
We therefore conducted a community-based cross-sectional survey in Urumqi, Xinjiang, to estimate the survey-based burden of self-reported physician-diagnosed KOA, past-year knee pain, and symptomatic KOA; describe pain-related functional impact and health-care seeking; and examine demographic, clinical, lifestyle, and occupational factors associated with these outcomes. The study was designed to provide locally relevant descriptive and associational evidence, rather than to establish causal effects.
2. Methods
2.1. Study design and participants
We conducted a cross-sectional, community-based survey in Urumqi City, Xinjiang, China, from June to August 2023. Adults aged 18 years or older who had resided in the city for at least six months were eligible. Participants completed an electronic questionnaire distributed via WeChat and community outreach. Each record was linked to a unique mobile phone number to prevent duplicate submissions. Records with duplicate numbers, ages outside 18–100 years, or missing core outcome data were excluded (Figure 1).
Figure 1.
Study population and analytical sample construction. Among the 76 participants excluded after age screening, mutually exclusive sequential reasons were invalid or missing height (n = 25), invalid or missing weight among those with valid height (n = 9), and BMI outside 15–45 kg/m2 despite valid height and weight (n = 42). No participant was excluded for missing physician-diagnosed KOA or past-year knee-pain status.
The study protocol was approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (approval number 230306-112). The present survey constituted the cross-sectional epidemiological assessment component of a broader ethics-approved project on KOA management; the submitted ethics materials included the community questionnaire, recruitment procedures, and consent process. No intervention or follow-up treatment data were used in the present analysis. All participants provided written electronic informed consent, and data were anonymised before analysis. The complete questionnaire is provided as Supplementary File S1, and reporting follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement (7).
2.2. Knee osteoarthritis and knee pain assessment
Primary outcomes were self-reported physician-diagnosed knee osteoarthritis and past-year knee pain. Physician-diagnosed knee osteoarthritis was defined by an affirmative response to the questionnaire item, “Have you ever been diagnosed with knee osteoarthritis by a doctor?” Past-year knee pain was defined by an affirmative response to “Have you had knee pain during the past 12 months?”; the questionnaire did not require a minimum frequency, duration, or severity threshold. Symptomatic knee osteoarthritis was defined as the coexistence of these two self-reported outcomes. Among participants reporting a history of knee pain, pain-related functional impact was assessed directly with three binary items asking whether knee pain affected sleep, affected work or other activities, or prompted medical help-seeking. Two further study-specific binary items assessed knee pain after exercise and in cold weather during the past 12 months. Because these outcomes were not confirmed by imaging, clinical examination, or medical records, they were treated as survey-based measures rather than clinically verified diagnoses (8).
2.3. Covariates
Self-reported sociodemographic variables included age, sex, ethnicity, residence, education, occupation, and monthly income. Monthly income referred to the participant’s own self-reported monthly income in Chinese yuan and was grouped as <3,000, 3,000–4,999, 5,000–7,999, and ≥8,000 CNY. Self-reported height and weight were used to calculate BMI (kg/m2). BMI was categorised according to the Working Group on Obesity in China criteria as underweight (<18.5 kg/m2), normal weight (18.5–23.9 kg/m2), overweight (24.0–27.9 kg/m2), or obesity (≥28.0 kg/m2) (9). Smoking status, weekly exercise, and routine regularity were obtained from separate questionnaire items.
Occupational mechanical load was represented by a study-specific score based on four current-work items: prolonged standing or walking, frequent knee bending or twisting, repeated squatting or stair climbing, and handling heavy loads. Each affirmative response was assigned 1 point and each negative response 0 points; item scores were summed to give a total from 0 to 4. Scores were analysed as 0, 1, 2, and 3–4, with the two upper categories combined to avoid sparse cells. This score was not a previously validated instrument and reflected concurrent work requirements rather than cumulative lifetime exposure. Sleep disorder and irregular routine were assessed as separate binary self-reports using the questions “Do you have a sleep disorder?” and “Is your daily routine regular?”, respectively. The metabolic comorbidity count comprised self-reported hypertension, diabetes, hyperlipidaemia, and hyperuricaemia and was categorised as 0, 1, or ≥2 conditions. Previous knee injury was self-reported. First-degree family history of arthritis was defined as any reported arthritis in a participant’s father, mother, or full sibling.
2.4. Statistical analysis
Analyses were conducted using R (version 4.3.2). Inspection of histograms and quantile plots indicated that age and BMI were not normally distributed; continuous variables were therefore summarised as medians with IQRs and compared with Wilcoxon rank-sum tests. Categorical variables were summarised as counts and percentages and compared with chi-square tests. Prevalence estimates were accompanied by Wilson score 95% CIs because these intervals have more reliable coverage than normal-approximation intervals, particularly for subgroup estimates with smaller denominators (10).
To obtain directly interpretable prevalence ratios, modified Poisson regression with robust error variance was selected a priori for the binary outcomes. This approach avoids the potential overstatement associated with odds ratios for common outcomes and is less prone to boundary-related convergence problems than log-binomial regression (11, 12). Exploratory log-binomial models using the same adjustment sets were attempted, but did not converge reliably for all three outcomes; modified Poisson models were therefore retained. The main factors of interest were BMI category, sleep disorder, occupational mechanical load score, previous knee injury, metabolic comorbidity count, and first-degree family history of arthritis. Models additionally adjusted a priori for age, sex, ethnicity, residence, education, monthly income, smoking status, weekly exercise, and routine regularity on the basis of epidemiological relevance rather than univariable statistical significance. The same adjustment set was used for physician-diagnosed KOA, past-year knee pain, and symptomatic KOA. BMI, metabolic comorbidity, and sleep disorder were retained because they represented distinct anthropometric, cardiometabolic, and symptom domains; potential multicollinearity was assessed using variance inflation factors. The three primary models included 26 covariate parameters and 236, 562, and 212 outcome events, respectively, corresponding to 9.1, 21.6, and 8.2 events per parameter. These ratios were used descriptively rather than as fixed adequacy thresholds; all models converged, and estimates for symptomatic KOA were interpreted with particular attention to CI width (13).
For participants reporting past-year knee pain, functional-impact outcomes were examined in secondary, exploratory modified Poisson models using a prespecified parsimonious set of age, sex, BMI category, sleep disorder, occupational load score 3–4, previous knee injury, metabolic comorbidity, and first-degree family history of arthritis. Primary estimates are presented with unadjusted two-sided p values because the exposure–outcome comparisons were prespecified; however, a Benjamini–Hochberg false-discovery-rate sensitivity analysis was applied across the 45 prespecified exposure–outcome comparisons, and the smaller functional-impact analyses were interpreted as exploratory (14). Sensitivity analyses also excluded rapid responses (<180 s and <300 s), restricted the sample to participants with likely Urumqi residency according to IP address or residential address, excluded records with logical conflicts, repeated the analyses among all adults aged 18 years or older, and restricted the primary population to adults aged 45 years or older.
All variables entered into the primary regression models were complete in the main analytical sample; therefore, the primary analyses used complete-case data, no multiple imputation was performed, and a complete-case versus multiple-imputation comparison was not applicable. Conditional pain-impact items had item-specific denominators because of questionnaire skip logic and item nonresponse and were analysed among participants with non-missing responses to the corresponding item. All tests were two-sided. Effect estimates and 95% CIs were prioritised over dichotomous interpretation of the conventional p < 0.05 threshold.
3. Results
3.1. Study population
A total of 4,117 questionnaire records were submitted. After exclusion of 116 duplicate records identified from valid mobile phone numbers, 4,001 records remained. Of these, 45 were excluded because age was outside 18–100 years. Among the remaining 3,956 participants, 76 were excluded using mutually exclusive sequential criteria: 25 had invalid or missing height, nine had invalid or missing weight after removal of those with invalid height, and 42 had a calculated BMI outside 15–45 kg/m2 despite valid height and weight. No participant was excluded because physician-diagnosed KOA or past-year knee-pain status was missing. The resulting 3,880 participants had valid anthropometric information and core outcomes; 1925 adults aged 40 years or older formed the primary analytical sample. Prespecified sensitivity analyses excluded responses completed in less than 180 s (n = 1834), excluded responses completed in less than 300 s (n = 1,345), and restricted the sample to a likely Urumqi source according to IP address or residential address (n = 1730; Figure 1; Supplementary Table S1). No values were missing for variables included in the primary regression models; conditional pain-impact items had item-specific denominators because of skip logic and item nonresponse.
3.2. Baseline characteristics
Among the 1925 participants included in the main analysis, the median age was 55.0 years (IQR 48.0–63.0), and 1,118 (58.1%) were men. Overall, 1,406 (73.0%) participants were Han Chinese and 519 (27.0%) were classified as other ethnic groups. Most participants lived in urban areas (1,655 [86.0%]). The median BMI was 23.9 kg/m2 (IQR 21.6–26.4); 696 (36.2%) participants were overweight and 258 (13.4%) had obesity. Sleep disorder was reported by 454 (23.6%) participants, 423 (22.0%) had a high occupational mechanical load score of 3–4, 184 (9.6%) reported previous knee injury, and 342 (17.8%) reported a first-degree family history of arthritis (Table 1).
Table 1.
Characteristics of participants aged 40 years or older, by self-reported physician-diagnosed knee osteoarthritis.
| Characteristic | Overall (n = 1925) | No KOA (n = 1,689) | KOA (n = 236) | p value |
|---|---|---|---|---|
| Age, years | 55.0 (48.0–63.0) | 55.0 (48.0–62.0) | 58.0 (49.0–65.0) | 0.006 |
| Age group | ||||
| 40–49 | 621 (32.3%) | 553 (32.7%) | 68 (28.8%) | 0.232 |
| 50–59 | 651 (33.8%) | 577 (34.2%) | 74 (31.4%) | |
| 60–69 | 416 (21.6%) | 357 (21.1%) | 59 (25.0%) | |
| ≥70 | 237 (12.3%) | 202 (12.0%) | 35 (14.8%) | |
| Sex | ||||
| Male | 1,118 (58.1%) | 1014 (60.0%) | 104 (44.1%) | <0.001 |
| Female | 807 (41.9%) | 675 (40.0%) | 132 (55.9%) | |
| Ethnicity | ||||
| Han | 1,406 (73.0%) | 1250 (74.0%) | 156 (66.1%) | 0.013 |
| Other | 519 (27.0%) | 439 (26.0%) | 80 (33.9%) | |
| Residence | ||||
| Urban | 1,655 (86.0%) | 1456 (86.2%) | 199 (84.3%) | 0.316 |
| Suburban | 21 (1.1%) | 20 (1.2%) | 1 (0.4%) | |
| Rural | 249 (12.9%) | 213 (12.6%) | 36 (15.3%) | |
| Education | ||||
| Primary or below | 307 (15.9%) | 258 (15.3%) | 49 (20.8%) | 0.004 |
| Junior/senior high | 913 (47.4%) | 824 (48.8%) | 89 (37.7%) | |
| College or above | 705 (36.6%) | 607 (35.9%) | 98 (41.5%) | |
| Monthly income, CNY | ||||
| <3,000 | 425 (22.1%) | 367 (21.7%) | 58 (24.6%) | 0.350 |
| 3,000–4,999 | 598 (31.1%) | 520 (30.8%) | 78 (33.1%) | |
| 5,000–7,999 | 645 (33.5%) | 578 (34.2%) | 67 (28.4%) | |
| ≥8,000 | 257 (13.4%) | 224 (13.3%) | 33 (14.0%) | |
| Occupation | ||||
| Manual/physical | 263 (13.7%) | 223 (13.2%) | 40 (16.9%) | 0.054 |
| Office/professional/student | 830 (43.1%) | 716 (42.4%) | 114 (48.3%) | |
| Other/unknown | 568 (29.5%) | 509 (30.1%) | 59 (25.0%) | |
| Self-employed/flexible | 203 (10.5%) | 187 (11.1%) | 16 (6.8%) | <0.001 |
| Unemployed | 61 (3.2%) | 54 (3.2%) | 7 (3.0%) | |
| BMI, kg/m2 | 23.9 (21.6–26.4) | 23.7 (21.5–26.3) | 25.0 (22.9–27.0) | |
| BMI category | ||||
| Underweight | 56 (2.9%) | 55 (3.3%) | 1 (0.4%) | <0.001 |
| Normal | 915 (47.5%) | 827 (49.0%) | 88 (37.3%) | |
| Overweight | 696 (36.2%) | 592 (35.1%) | 104 (44.1%) | |
| Obesity | 258 (13.4%) | 215 (12.7%) | 43 (18.2%) | |
| Smoking status | ||||
| Never | 1234 (64.1%) | 1061 (62.8%) | 173 (73.3%) | 0.006 |
| Former | 315 (16.4%) | 284 (16.8%) | 31 (13.1%) | |
| Current | 376 (19.5%) | 344 (20.4%) | 32 (13.6%) | |
| Weekly exercise | ||||
| ≥2.5 h/week | 281 (14.6%) | 257 (15.2%) | 24 (10.2%) | 0.007 |
| 1–2.5 h/week | 844 (43.8%) | 751 (44.5%) | 93 (39.4%) | |
| ≤1 h/week | 800 (41.6%) | 681 (40.3%) | 119 (50.4%) | |
| Sleep disorder | ||||
| Yes | 454 (23.6%) | 337 (20.0%) | 117 (49.6%) | <0.001 |
| No | 1471 (76.4%) | 1352 (80.0%) | 119 (50.4%) | |
| Irregular routine | ||||
| Yes | 617 (32.1%) | 557 (33.0%) | 60 (25.4%) | 0.024 |
| No | 1308 (67.9%) | 1132 (67.0%) | 176 (74.6%) | |
| Occupational load score | ||||
| 0 | 1153 (59.9%) | 1043 (61.8%) | 110 (46.6%) | <0.001 |
| 1 | 239 (12.4%) | 202 (12.0%) | 37 (15.7%) | |
| 2 | 110 (5.7%) | 94 (5.6%) | 16 (6.8%) | |
| 3–4 | 423 (22.0%) | 350 (20.7%) | 73 (30.9%) | |
| Prior knee injury | ||||
| Yes | 184 (9.6%) | 100 (5.9%) | 84 (35.6%) | <0.001 |
| No | 1741 (90.4%) | 1589 (94.1%) | 152 (64.4%) | |
| Metabolic comorbidity count | ||||
| 0 | 966 (50.2%) | 863 (51.1%) | 103 (43.6%) | 0.001 |
| 1 | 677 (35.2%) | 597 (35.3%) | 80 (33.9%) | |
| ≥2 | 282 (14.6%) | 229 (13.6%) | 53 (22.5%) | |
| First-degree family history of arthritis | ||||
| Yes | 342 (17.8%) | 253 (15.0%) | 89 (37.7%) | <0.001 |
| No | 1583 (82.2%) | 1436 (85.0%) | 147 (62.3%) | |
Data are median (IQR) or n (%). Age and BMI were non-normally distributed and were compared with Wilcoxon rank-sum tests; categorical variables were compared with chi-square tests. No values were missing for the variables shown, and This table presents the original complete-case data. KOA = knee osteoarthritis.
Participants with self-reported physician-diagnosed knee osteoarthritis differed from those without knee osteoarthritis in several clinically relevant characteristics. They were older (median age 58.0 years [IQR 49.0–65.0] vs. 55.0 years [48.0–62.0]; p = 0.006), more often female (132 [55.9%] of 236 vs. 675 [40.0%] of 1,689; p < 0.001), and had a higher median BMI (25.0 kg/m2 [IQR 22.9–27.0] vs. 23.7 kg/m2 [21.5–26.3]; p < 0.001). They were also more likely to report sleep disorder (117 [49.6%] vs. 337 [20.0%]), previous knee injury (84 [35.6%] vs. 100 [5.9%]), two or more metabolic comorbidities (53 [22.5%] vs. 229 [13.6%]), and a first-degree family history of arthritis (89 [37.7%] vs. 253 [15.0%]; all p < 0.001; Table 1).
3.3. Prevalence of knee osteoarthritis and knee pain
In the primary analytical sample, 236 participants reported physician-diagnosed knee osteoarthritis, corresponding to a prevalence of 12.3% (95% CI 10.9–13.8). Past-year knee pain was reported by 562 participants (29.2, 95% CI 27.2–31.3). Symptomatic knee osteoarthritis, defined as physician-diagnosed knee osteoarthritis with concurrent past-year knee pain, was present in 212 participants (11.0%, 9.7–12.5; Table 2).
Table 2.
Burden of knee osteoarthritis, knee pain, and pain-related functional impact.
| Outcome | Denominator | n/N | Prevalence (95% CI) |
|---|---|---|---|
| Self-reported physician-diagnosed knee osteoarthritis | Main sample (n = 1925) | 236/1925 | 12.3% (10.9–13.8) |
| Past-year knee pain | Main sample (n = 1925) | 562/1925 | 29.2% (27.2–31.3) |
| Symptomatic knee osteoarthritis (diagnosed KOA + past-year knee pain) | Main sample (n = 1925) | 212/1925 | 11.0% (9.7–12.5) |
| Knee pain affecting sleep | Participants with past-year knee pain (n = 562) | 214/525 | 40.8% (36.6–45.0) |
| Knee pain affecting work or activities | Participants with past-year knee pain (n = 562) | 286/525 | 54.5% (50.2–58.7) |
| Sought medical help for knee pain | Participants with past-year knee pain (n = 562) | 320/525 | 61.0% (56.7–65.0) |
| Exercise-related knee pain in past year | Participants with past-year knee pain (n = 562) | 363/385 | 94.3% (91.5–96.2) |
| Cold-weather knee pain in past year | Participants with past-year knee pain (n = 562) | 332/345 | 96.2% (93.7–97.8) |
| Any sleep or work/activity impact among those with past-year knee pain | Participants with past-year knee pain (n = 562) | 309/562 | 55.0% (50.8–59.0) |
Prevalence estimates are shown with Wilson 95% CIs. Of 562 participants with past-year knee pain, 37 reported no previous history of knee pain and therefore did not receive the conditional sleep, work/activity, and medical-help items, leaving n = 525. The exercise-related pain item was conditional on a history of exercise-related knee pain (n = 385), and the cold-weather item was conditional on a history of cold-weather knee pain (n = 345). The n/N values therefore reflect questionnaire skip logic and item-specific nonresponse rather than longitudinal follow-up.
The burden of all three outcomes rose with age. The prevalence of physician-diagnosed knee osteoarthritis increased from 11.0% (95% CI 8.7–13.7) among participants aged 40–49 years to 14.8% (10.8–19.8) among those aged 70 years or older. Over the same age range, past-year knee pain increased from 27.4% (24.0–31.0) to 33.3% (27.6–39.6), and symptomatic knee osteoarthritis increased from 9.7% (7.6–12.2) to 13.9% (10.1–18.9; Figure 2; Supplementary Table S2).
Figure 2.
Age-specific prevalence of self-reported physician-diagnosed knee osteoarthritis, past-year knee pain, and symptomatic knee osteoarthritis.
Women had a consistently higher burden than men. Past-year knee pain was reported by 342 (42.3, 95% CI 38.9–45.7) of 807 women and 221 (19.8%, 17.5–22.2) of 1,118 men. Physician-diagnosed knee osteoarthritis was also more common among women than men (16.4% [14.0–19.1] vs. 9.3% [7.7–11.1]), as was symptomatic knee osteoarthritis (15.1% [12.8–17.8] vs. 8.1% [6.6–9.8]; Supplementary Table S2 and Supplementary Figure S1).
3.4. Functional impact and health-care use among participants with past-year knee pain
Past-year knee pain was frequently accompanied by sleep disturbance, activity limitation, and health-care use. Among participants with past-year knee pain and non-missing responses to the corresponding conditional questionnaire items, 214 of 525 reported that knee pain affected sleep (40.8, 95% CI 36.6–45.0), 286 of 525 reported that knee pain affected work or daily activities (54.5%, 50.2–58.7), and 320 of 525 had sought medical care for knee pain (61.0%, 56.7–65.0). Among participants who completed the relevant conditional items, 363 of 385 reported exercise-related knee pain in the past year (94.3%, 91.5–96.2), and 332 of 345 reported cold-weather knee pain (96.2%, 93.7–97.8). Overall, 309 of 562 participants with past-year knee pain reported either sleep disturbance or work/activity limitation attributable to knee pain (55.0%, 50.8–59.0; Table 2; Figure 3).
Figure 3.
Pain-related functional impact and health-care seeking among participants with past-year knee pain.
3.5. Burden by BMI and occupational mechanical load
Knee outcomes varied across BMI categories. Physician-diagnosed knee osteoarthritis was reported by 88 (9.6, 95% CI 7.9–11.7) of 915 participants with normal BMI, 104 (14.9%, 12.5–17.8) of 696 participants with overweight, and 43 (16.7%, 12.6–21.7) of 258 participants with obesity. A similar pattern was seen for symptomatic knee osteoarthritis, which increased from 8.4% (6.8–10.4) in participants with normal BMI to 13.6% (11.3–16.4) in those with overweight and 15.5% (11.6–20.4) in those with obesity (Supplementary Table S2; Supplementary Figure S2).
The prevalence of knee outcomes also differed by occupational mechanical load. Physician-diagnosed knee osteoarthritis was lowest among participants with an occupational mechanical load score of 0 (9.5, 95% CI 8.0–11.4) and highest among those with a score of 3–4 (17.3%, 14.0–21.1). Symptomatic knee osteoarthritis increased from 8.7% (7.2–10.4) in the score 0 group to 14.9% (11.8–18.6) in the score 3–4 group. Past-year knee pain was 28.1% (25.6–30.8) among participants with a score of 0, 36.0% (30.2–42.2) among those with a score of 1, 34.5% (26.3–43.8) among those with a score of 2, and 27.0% (22.9–31.4) among those with a score of 3–4 (Figure 4; Supplementary Table S2).
Figure 4.
Burden of knee osteoarthritis and knee pain by occupational mechanical load score.
3.6. Factors associated with knee osteoarthritis and past-year knee pain
In the fully adjusted modified Poisson models, age showed consistent associations with all three outcomes (Table 3; Figure 5). Each 10-year increase in age was associated with physician-diagnosed KOA (aPR 1.27, 95% CI 1.12–1.44), past-year knee pain (1.19, 1.11–1.27), and symptomatic KOA (1.29, 1.12–1.48; all p < 0.001). Female sex was clearly associated with past-year knee pain (aPR 1.73, 95% CI 1.43–2.08; p < 0.001). Estimates for physician-diagnosed KOA (1.36, 1.00–1.86; p = 0.050) and symptomatic KOA (1.40, 1.01–1.94; p = 0.046) were less precise, with lower confidence limits close to 1. Variance inflation factors were all below 2.4, providing no indication of problematic statistical multicollinearity.
Table 3.
Adjusted associations of selected demographic, clinical, lifestyle, and occupational factors with knee osteoarthritis and knee pain.
| Predictor | Physician-diagnosed KOA | Past-year knee pain | Symptomatic KOA |
|---|---|---|---|
| aPR (95% CI); p value | aPR (95% CI); p value | aPR (95% CI); p value | |
| Age (per 10 years) | 1.27 (1.12–1.44); p < 0.001 | 1.19 (1.11–1.27); p < 0.001 | 1.29 (1.12–1.48); p < 0.001 |
| Female sex | 1.36 (1.00–1.86); p = 0.050 | 1.73 (1.43–2.08); p < 0.001 | 1.40 (1.01–1.94); p = 0.046 |
| Overweight | 1.47 (1.14–1.90); p = 0.003 | 1.15 (1.00–1.33); p = 0.049 | 1.54 (1.17–2.02); p = 0.002 |
| Obesity | 1.39 (1.00–1.92); p = 0.050 | 1.06 (0.87–1.29); p = 0.573 | 1.43 (1.01–2.02); p = 0.044 |
| Exercise 1–2.5 h/week | 1.15 (0.78–1.71); p = 0.484 | 0.95 (0.77–1.16); p = 0.591 | 1.05 (0.70–1.57); p = 0.821 |
| Exercise ≤1 h/week | 1.26 (0.85–1.86); p = 0.250 | 0.98 (0.81–1.20); p = 0.849 | 1.17 (0.79–1.75); p = 0.435 |
| Sleep disorder | 2.04 (1.59–2.62); p < 0.001 | 1.81 (1.58–2.08); p < 0.001 | 2.32 (1.77–3.03); p < 0.001 |
| Irregular routine | 0.80 (0.61–1.05); p = 0.112 | 1.02 (0.89–1.17); p = 0.766 | 0.84 (0.63–1.12); p = 0.228 |
| Occupational load score 1 | 1.29 (0.91–1.83); p = 0.153 | 0.99 (0.82–1.20); p = 0.929 | 1.27 (0.87–1.84); p = 0.215 |
| Occupational load score 2 | 1.29 (0.77–2.15); p = 0.339 | 0.92 (0.70–1.20); p = 0.524 | 1.34 (0.79–2.26); p = 0.274 |
| Occupational load score 3–4 | 1.48 (1.09–2.03); p = 0.013 | 0.93 (0.78–1.12); p = 0.446 | 1.39 (1.00–1.91); p = 0.048 |
| Prior knee injury | 3.08 (2.37–4.00); p < 0.001 | 2.01 (1.72–2.34); p < 0.001 | 2.97 (2.25–3.91); p < 0.001 |
| One metabolic comorbidity | 1.01 (0.78–1.31); p = 0.963 | 1.01 (0.87–1.17); p = 0.881 | 1.03 (0.79–1.36); p = 0.811 |
| ≥2 metabolic comorbidities | 1.36 (0.99–1.87); p = 0.059 | 1.11 (0.92–1.34); p = 0.292 | 1.37 (0.98–1.92); p = 0.069 |
| First-degree family history of arthritis | 1.81 (1.39–2.37); p < 0.001 | 1.66 (1.44–1.92); p < 0.001 | 1.94 (1.47–2.57); p < 0.001 |
Adjusted prevalence ratios (aPRs) were estimated using modified Poisson regression with robust standard errors. The same fully adjusted model was used for each outcome and included age, sex, ethnicity (Han or other), residence, education, personal monthly income, BMI category, smoking, weekly exercise, sleep disorder, routine regularity, occupational mechanical load, previous knee injury, metabolic comorbidity count, and first-degree family history of arthritis. Reference categories are defined in the methods. p values are unadjusted; a Benjamini–Hochberg false-discovery-rate sensitivity analysis was undertaken across the 45 displayed comparisons. KOA = knee os.
Figure 5.
Visual summary of selected prespecified factors from the fully adjusted models for self-reported physician-diagnosed knee osteoarthritis and past-year knee pain. The complete model results are reported in Table 3; the figure is retained to facilitate comparison of effect magnitude and precision. Error bars show 95% CIs. aPR = adjusted prevalence ratio; KOA = knee osteoarthritis.
Compared with normal BMI, overweight was associated with physician-diagnosed KOA (aPR 1.47, 95% CI 1.14–1.90; p = 0.003) and symptomatic KOA (1.54, 1.17–2.02; p = 0.002). The estimate for past-year knee pain was smaller and imprecise (1.15, 1.00–1.33; p = 0.049). For obesity, estimates for physician-diagnosed KOA (1.39, 1.00–1.92; p = 0.050) and symptomatic KOA (1.43, 1.01–2.02; p = 0.044) also had lower confidence limits close to 1, and there was no clear association with past-year knee pain (1.06, 0.87–1.29; p = 0.573; Table 3).
Sleep disorder showed comparatively large and precise associations with physician-diagnosed KOA (aPR 2.04, 95% CI 1.59–2.62), past-year knee pain (1.81, 1.58–2.08), and symptomatic KOA (2.32, 1.77–3.03; all p < 0.001). An occupational mechanical load score of 3–4 was associated with physician-diagnosed KOA (aPR 1.48, 95% CI 1.09–2.03; p = 0.013). The estimate for symptomatic KOA was weaker and borderline (1.39, 1.00–1.91; p = 0.048), and there was no association with past-year knee pain (0.93, 0.78–1.12; p = 0.446; Table 3; Figure 5).
Previous knee injury and first-degree family history of arthritis showed the most consistent associations across outcomes. Previous knee injury was associated with physician-diagnosed KOA (aPR 3.08, 95% CI 2.37–4.00), past-year knee pain (2.01, 1.72–2.34), and symptomatic KOA (2.97, 2.25–3.91; all p < 0.001). Corresponding estimates for first-degree family history were 1.81 (1.39–2.37), 1.66 (1.44–1.92), and 1.94 (1.47–2.57; all p < 0.001). Estimates for ≥2 metabolic comorbidities were compatible with both no association and modestly increased prevalence for physician-diagnosed KOA (1.36, 0.99–1.87; p = 0.059) and symptomatic KOA (1.37, 0.98–1.92; p = 0.069), and no clear association was observed for past-year knee pain (1.11, 0.92–1.34; p = 0.292; Table 3; Figure 5).
3.7. Factors associated with functional burden among participants with past-year knee pain
Among participants with past-year knee pain, several factors were associated with greater functional burden. Previous knee injury was associated with knee pain affecting sleep (aPR 1.57, 95% CI 1.27–1.93; p < 0.001), affecting work or daily activities (1.30, 1.11–1.52; p = 0.001), medical care seeking (1.39, 1.22–1.58; p < 0.001), and any sleep or work/activity impact (1.28, 1.10–1.48; p = 0.001). A high occupational mechanical load score of 3–4 was associated with sleep impact (aPR 1.42, 95% CI 1.08–1.85; p = 0.011), work or activity impact (1.44, 1.18–1.75; p < 0.001), medical care seeking (1.21, 1.02–1.43; p = 0.025), and any sleep or work/activity impact (1.41, 1.17–1.71; p < 0.001; Supplementary Table S3).
Higher BMI was also related to functional consequences of knee pain. Obesity was associated with work or activity impact (aPR 1.45, 95% CI 1.15–1.83; p = 0.002), medical care seeking (1.27, 1.07–1.52; p = 0.008), and any sleep or work/activity impact (1.37, 1.10–1.70; p = 0.005). Overweight showed similar associations with work or activity impact (aPR 1.40, 95% CI 1.19–1.66; p < 0.001), medical care seeking (1.17, 1.01–1.36; p = 0.042), and any sleep or work/activity impact (1.37, 1.17–1.61; p < 0.001). Sleep disorder was associated with sleep impact (aPR 1.43, 95% CI 1.17–1.76; p < 0.001), work or activity impact (1.22, 1.05–1.42; p = 0.011), and any sleep or work/activity impact (1.33, 1.15–1.54; p < 0.001). Two or more metabolic comorbidities were associated with all functional-impact outcomes, including sleep impact (aPR 1.46, 95% CI 1.12–1.89; p = 0.004), work or activity impact (1.44, 1.17–1.78; p < 0.001), medical care seeking (1.24, 1.04–1.48; p = 0.017), and any sleep or work/activity impact (1.35, 1.11–1.65; p = 0.003; Supplementary Table S3).
3.8. Sensitivity analyses
The principal patterns were similar across data-quality and population-restriction analyses. After exclusion of responses completed in less than 180 s, estimates for physician-diagnosed KOA remained similar for sleep disorder (aPR 1.97, 95% CI 1.54–2.53), occupational load score 3–4 (1.52, 1.11–2.09), and previous knee injury (3.02, 2.33–3.92). Results were also broadly unchanged after exclusion of responses completed in less than 300 s, restriction to a likely Urumqi source, exclusion of logical conflicts, and analysis of all adults aged ≥18 years (Supplementary Table S4). In the analysis restricted to participants aged ≥45 years (n = 1,636; 206 physician-diagnosed KOA, 480 past-year knee-pain, and 185 symptomatic-KOA events), estimates remained similar for previous knee injury (aPRs 3.12, 1.89, and 2.97, respectively), sleep disorder (2.14, 1.74, and 2.41), and first-degree family history of arthritis (1.74, 1.62, and 1.93). The false-discovery-rate sensitivity analysis supported the principal associations with age, sleep disorder, previous knee injury, family history, and overweight for physician-diagnosed and symptomatic KOA; several estimates with unadjusted p values near 0.05 did not remain below the 5% false-discovery-rate threshold. Exploratory log-binomial models did not converge reliably across all outcomes.
4. Discussion
In this community-based survey of adults aged 40 years or older in Urumqi, knee osteoarthritis and knee pain were common and were accompanied by substantial functional burden. Approximately one in eight participants reported physician-diagnosed knee osteoarthritis, nearly three in ten reported knee pain in the past year, and one in nine met our definition of symptomatic knee osteoarthritis. Among participants with past-year knee pain, more than 40% reported sleep disturbance, more than half reported interference with work or daily activities, and more than 60% had sought medical care. These findings indicate that knee osteoarthritis and knee pain in this setting are not only musculoskeletal diagnoses, but also common chronic health problems that affect sleep, mobility, work capacity, and use of primary-care services.
The prevalence observed in our study is broadly consistent with previous evidence from China and international studies, although comparisons should be made with caution because of differences in definitions, age structure, and sampling methods. In the China Health and Retirement Longitudinal Study, symptomatic knee osteoarthritis affected 8.1% of adults aged 45 years or older, with higher prevalence in women, older adults, less educated groups, and some western and southwestern regions (15). A meta-analysis of Chinese studies also showed a substantial burden of symptomatic knee osteoarthritis, particularly among women and older individuals (16). Our estimate of symptomatic knee osteoarthritis was somewhat higher than that reported in some national datasets, but it remains plausible for a middle-aged and older community sample in northwest China. At the global level, the GBD 2021 analysis estimated that 595 million people were living with osteoarthritis in 2020, with knee osteoarthritis the most common anatomical site and a marked increase projected by 2050 (17). Analyses covering 204 countries and territories similarly showed that the prevalence, incidence, and years lived with disability attributable to knee osteoarthritis have increased over recent decades, with high BMI contributing an increasing share of disease burden (4). Our findings therefore reflect a local manifestation of a broader transition: musculoskeletal disability is becoming an increasingly important component of chronic disease burden in ageing and urbanising populations.
A key feature of this study is the distinction between self-reported physician-diagnosed knee osteoarthritis, past-year knee pain, and symptomatic knee osteoarthritis. The symptom burden was greater than the burden captured by physician-diagnosis history alone. This distinction is relevant to clinical and public health interpretation, but it also highlights the limitations of survey-based definitions. Participants with knee pain who had not sought care or received a diagnosis were classified as not having physician-diagnosed knee osteoarthritis, whereas past-year knee pain could reflect conditions other than knee osteoarthritis. Diagnostic history may also vary with access to care, health awareness, and recall. The differences between these outcomes should therefore not be interpreted as evidence of an undiagnosed-disease gap without clinical confirmation. Previous studies indicate that pain and high-impact symptoms are closely related to physical function, disability, and health-care use (5), while the accuracy of self-reported arthritis is acceptable but imperfect in large epidemiological studies (8). Our findings support reporting diagnostic history and symptom-related impact as complementary, rather than interchangeable, measures.
The high proportions of participants reporting exercise-related and cold-weather knee pain should be interpreted descriptively. Exercise-related pain is not evidence that physical activity is harmful; it may instead identify people who need individualised guidance on safe, progressive activity. Exercise is recommended in major osteoarthritis guidelines and improves pain and function in intervention studies (18–20). In Urumqi, long cold seasons might influence symptom perception and outdoor activity, but the cross-sectional questionnaire cannot establish that cold exposure caused knee pain or activity avoidance. These observations could inform the design of future prospective studies and locally adapted symptom-management programmes, including indoor exercise options and seasonal advice, but they do not demonstrate the effectiveness of such strategies.
The associations between age, sex, and knee outcomes were consistent with established epidemiological evidence. Older age was associated with physician-diagnosed knee osteoarthritis, past-year knee pain, and symptomatic knee osteoarthritis. Women had a substantially higher prevalence of past-year knee pain and symptomatic knee osteoarthritis than men. Similar patterns have been reported in global and Chinese studies, in which women and older adults carry a disproportionate burden of knee osteoarthritis (4, 15–17). The sex difference is likely to reflect a combination of biological, biomechanical, and social factors, including differences in body composition, muscle strength, pain perception, health-care seeking, lifetime physical workload, and post-menopausal changes. Because our study did not measure hormonal status, joint alignment, or muscle strength, these mechanisms could not be examined directly. Nevertheless, the findings support the need for sex-sensitive prevention and management strategies, especially for middle-aged and older women.
Higher BMI was associated with physician-diagnosed and symptomatic knee osteoarthritis in this study. Prospective meta-analyses and cohort studies have reported dose–response and longitudinal associations between adiposity and knee osteoarthritis (21, 22), and mechanical loading and metabolic inflammation are plausible mechanisms described in previous work (23). However, our cross-sectional data cannot determine whether excess body weight preceded knee symptoms, whether pain-related activity restriction contributed to weight gain, or whether both reflected shared determinants. Metabolic comorbidity showed weaker and less precise associations than BMI. Because BMI and metabolic conditions are interrelated, the mutually adjusted coefficients should be interpreted as conditional associations rather than independent causal effects. Although variance inflation factors did not indicate problematic collinearity, residual confounding, mediation, and common-cause pathways cannot be separated in this design.
Previous knee injury had the largest adjusted prevalence ratios for all three knee outcomes. This pattern is consistent with longitudinal and meta-analytic evidence linking traumatic knee injury with later osteoarthritis (24, 25). Nevertheless, the survey did not capture the timing, anatomical diagnosis, severity, or treatment of injury, and participants with current knee problems might recall previous injuries more readily. The observed associations therefore support previous knee injury as a clinically relevant marker, but do not quantify the causal effect of injury or the benefit of any specific prevention or rehabilitation strategy.
A higher study-specific occupational mechanical load score was associated with physician-diagnosed and symptomatic knee osteoarthritis, and, among participants with knee pain, with several functional-impact outcomes. Previous studies have linked kneeling, squatting, climbing, lifting, and prolonged standing with knee osteoarthritis (26). However, our score measured current rather than lifetime work requirements and was not a validated cumulative exposure instrument. Its lack of an independent association with past-year knee pain, together with the non-monotonic descriptive pattern across score categories, argues against a simple dose–response interpretation. Reverse selection is also possible because people with more severe symptoms may reduce demanding tasks or leave physically intensive work. The results should therefore be interpreted as cross-sectional occupational correlates, not as evidence that the measured work activities caused the outcomes.
Sleep disorder was associated with all three knee outcomes and with several functional-impact measures. This relationship is likely to be complex and potentially bidirectional: pain can disrupt sleep, whereas poor sleep can heighten pain sensitivity, fatigue, and distress. Systematic reviews and longitudinal studies support links between sleep and osteoarthritis symptoms (27–29), but the single self-reported sleep item and cross-sectional timing in our study do not permit temporal ordering. Sleep disorder should therefore be viewed as a marker of greater concurrent symptom burden and a candidate for further longitudinal assessment, rather than as a demonstrated cause of knee osteoarthritis or knee pain.
The findings have potential implications for community assessment, but they should not be read as evidence that modifying the associated factors identified here will necessarily reduce knee osteoarthritis or knee pain. Asking about past-year knee pain, pain-related sleep or activity impact, previous knee injury, body size, sleep problems, and occupational demands could help characterise symptom burden and identify groups for further clinical assessment. Education, exercise, and weight management are recommended as core non-surgical treatments on the basis of external guideline and intervention evidence (18, 19), not on the basis of causal effects estimated in this survey. The high proportion of participants who had sought medical care suggests a need to examine whether existing primary-care pathways provide timely assessment, appropriate exercise advice, and referral for persistent or disabling symptoms.
This study has several strengths. It used a large community-recruited sample from Urumqi and captured both diagnostic history and symptom-related burden, providing a broader description than prevalence of physician diagnosis alone. The questionnaire included anthropometric, sleep, injury, comorbidity, family-history, and occupational characteristics, and the analyses distinguished physician-diagnosed knee osteoarthritis, past-year knee pain, symptomatic knee osteoarthritis, and functional outcomes among participants with pain. Duplicate removal, prespecified data-quality restrictions, sensitivity analyses, and explicit collinearity assessment supported the consistency and transparency of the findings.
Several limitations should be acknowledged. First, the cross-sectional design precludes causal or temporal inference. In particular, the direction of the associations involving sleep, BMI, metabolic comorbidity, physical activity, and occupational load cannot be established, and these variables may act as confounders, intermediates, or consequences of pain. Although variance inflation factors were low, statistical collinearity diagnostics do not resolve these causal relationships, and residual confounding by depression, pain catastrophising, medication use, physical capacity, and other unmeasured characteristics remains possible. Second, KOA was based on self-reported physician-diagnosis history and was not confirmed by radiography, clinical examination, or medical records. This definition might underestimate prevalence among symptomatic participants who had not sought care or received a diagnosis, but misunderstanding or inaccurate recall could also lead to over-reporting. Past-year knee pain may be more sensitive but less specific because it can arise from conditions other than osteoarthritis. If misclassification was unrelated to the associated factors, prevalence-ratio estimates would generally be expected to move towards the null; however, differential diagnosis or reporting—for example, greater clinical contact among participants with previous injury, obesity, or occupational symptoms—could bias estimates in either direction. Third, height, weight, injury, comorbidities, work demands, sleep, and functional impact were self-reported and are subject to recall and reporting error. The questionnaire did not collect analgesic use, topical treatment, injections, or other knee-pain therapies, which limited interpretation of symptom severity and health-care seeking. Fourth, the primary and secondary analyses involved multiple comparisons. Although the main exposure–outcome comparisons were prespecified and a false-discovery-rate sensitivity analysis was added, findings with confidence intervals close to 1 should remain hypothesis-generating. The primary models converged with 212–562 events, but the symptomatic-KOA model and conditional functional-impact analyses had lower information per parameter; their estimates should therefore be interpreted with attention to precision rather than p values alone. Fifth, online and community recruitment may have over-represented urban, educated, digitally connected, or health-conscious residents; the results are survey-based estimates and should not be interpreted as population-representative prevalence for all adults in Urumqi. Finally, the survey did not measure pain intensity, radiographic severity, mood symptoms, knee alignment, muscle strength, inflammatory markers, objective physical activity, or lifetime occupational exposure, and ethnicity was grouped as Han or other, limiting more detailed assessment of population heterogeneity.
5. Conclusion
Knee osteoarthritis and knee pain were common among community-recruited adults aged 40 years or older in Urumqi and were frequently accompanied by sleep disturbance, activity limitation, and health-care seeking. Previous knee injury, sleep disorder, higher BMI, family history of arthritis, and occupational mechanical load were associated with one or more knee outcomes, but these cross-sectional estimates do not establish temporal sequence or causality. The findings provide locally relevant survey evidence for community assessment and hypothesis generation; longitudinal and intervention studies are needed to determine whether addressing these factors can reduce symptoms or disability.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This study was supported by the “Tianshan Talent” High-Level Medical and Health Training Program (TSYC202301B091).
Footnotes
Edited by: Shengfa Li, Chengdu Third People's Hospital, China
Reviewed by: Rodrigo Gallardo-Rodríguez, Universidad Andrés Bello, Chile
Qianying Cai, Shanghai Jiao Tong University, China
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving humans were approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (approval number: 230306-112). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
LS: Data curation, Formal analysis, Funding acquisition, Writing – original draft. SL: Data curation, Formal analysis, Writing – original draft. JZ: Conceptualization, Writing – original draft. GH: Conceptualization, Writing – original draft. AA: Data curation, Writing – review & editing. XL: Data curation, Writing – review & editing. HZ: Conceptualization, Writing – review & editing. PM: Conceptualization, Investigation, Supervision, Writing – original draft, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2026.1903768/full#supplementary-material
References
- 1.Langworthy M, Dasa V, Spitzer AI. Knee osteoarthritis: disease burden, available treatments, and emerging options. Therap Adv Musculoskeletal Dis. (2024) 16:1759720X241273009. doi: 10.1177/1759720X241273009, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ren J-L, Yang J, Hu W. The global burden of osteoarthritis knee: a secondary data analysis of a population-based study. Clin Rheumatol. (2025) 44:1769–810. doi: 10.1007/s10067-025-07347-6, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Li E, Tan J, Xu K, Pan Y, Xu P. Global burden and socioeconomic impact of knee osteoarthritis: a comprehensive analysis. Front Med. (2024) 11:1323091. doi: 10.3389/fmed.2024.1323091, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Yang G, Wang J, Liu Y, Lu H, He L, Ma C, et al. Burden of knee osteoarthritis in 204 countries and territories, 1990–2019: results from the global burden of disease study 2019. Arthritis Care Res. (2023) 75:2489–500. doi: 10.1002/acr.25158, [DOI] [PubMed] [Google Scholar]
- 5.George SZ, Allen KD, Alvarez C, Arbeeva L, Callahan LF, Nelson AE, et al. Prevalence and factors associated with high-impact chronic pain in knee osteoarthritis: the Johnston county health study. J Pain. (2024) 25:104687. doi: 10.1016/j.jpain.2024.104687, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pavel RMS, Purza AL, Tit DM, Radu AF, Iovanovici DC, Vasileva D, et al. Functional burden and quality of life in hip and knee osteoarthritis: a cross-sectional study. Medicina. (2025) 61:1155. doi: 10.3390/medicina61071155, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. (2007) 370:1453–7. doi: 10.1016/S0140-6736(07)61602-X, [DOI] [PubMed] [Google Scholar]
- 8.Peeters GG, Alshurafa M, Schaap L, de Vet HC. Diagnostic accuracy of self-reported arthritis in the general adult population is acceptable. J Clin Epidemiol. (2015) 68:452–9. doi: 10.1016/j.jclinepi.2014.09.019, [DOI] [PubMed] [Google Scholar]
- 9.Zhou B. Predictive values of body mass index and waist circumference to risk factors of related diseases in Chinese adult population. Zhonghua liu xing bing xue za zhi= Zhonghua liuxingbingxue zazhi. (2002) 23:5–10. [PubMed] [Google Scholar]
- 10.Newcombe RG. Two-sided confidence intervals for the single proportion: comparison of seven methods. Stat Med. (1998) 17:857–72. doi: 10.1002/(SICI)1097-0258(19980430)17:8<857::AID-SIM777>3.0.CO;2-E, [DOI] [PubMed] [Google Scholar]
- 11.Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. (2004) 159:702–6. doi: 10.1093/aje/kwh090, [DOI] [PubMed] [Google Scholar]
- 12.Barros AJ, Hirakata VN. Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio. BMC Med Res Methodol. (2003) 3:21. doi: 10.1186/1471-2288-3-21, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and cox regression. Am J Epidemiol. (2007) 165:710–8. doi: 10.1093/aje/kwk052, [DOI] [PubMed] [Google Scholar]
- 14.Bender R, Lange S. Adjusting for multiple testing—when and how? J Clin Epidemiol. (2001) 54:343–9. doi: 10.1016/S0895-4356(00)00314-0, [DOI] [PubMed] [Google Scholar]
- 15.Tang X, Wang S, Zhan S, Niu J, Tao K, Zhang Y, et al. The prevalence of symptomatic knee osteoarthritis in China: results from the China health and retirement longitudinal study. Arthr Rheumatol. (2016) 68:648–53. doi: 10.1002/art.39465, [DOI] [PubMed] [Google Scholar]
- 16.Li D, Li S, Chen Q, Xie X. The prevalence of symptomatic knee osteoarthritis in relation to age, sex, area, region, and body mass index in China: a systematic review and meta-analysis. Front Med. (2020) 7:304. doi: 10.3389/fmed.2020.00304, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Steinmetz JD, Culbreth GT, Haile LM, Rafferty Q, Lo J, Fukutaki KG, et al. Global, regional, and national burden of osteoarthritis, 1990–2020 and projections to 2050: a systematic analysis for the global burden of disease study 2021. Lancet Rheumatol. (2023) 5:e508–22. doi: 10.1016/S2665-9913(23)00163-7, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kolasinski SL, Neogi T, Hochberg MC, Oatis C, Guyatt G, Block J, et al. 2019 American College of Rheumatology/Arthritis Foundation guideline for the management of osteoarthritis of the hand, hip, and knee. Arthr Rheumatol. (2020) 72:220–33. doi: 10.1002/art.41142, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bannuru RR, Osani M, Vaysbrot E, Osani MC, Vaysbrot EE, Arden NK, et al. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthr Cartil. (2019) 27:1578–89. doi: 10.1016/j.joca.2019.06.011 [DOI] [PubMed] [Google Scholar]
- 20.Mo L, Jiang B, Mei T, Zhou D. Exercise therapy for knee osteoarthritis: a systematic review and network meta-analysis. Orthop J Sports Med. (2023) 11:23259671231172773. doi: 10.1177/23259671231172773, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zheng H, Chen C. Body mass index and risk of knee osteoarthritis: systematic review and meta-analysis of prospective studies. BMJ Open. (2015) 5:e007568. doi: 10.1136/bmjopen-2014-007568, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Park D, Park Y-M, Ko S-H, Hyun KS, Choi YH, Min DU, et al. Association of general and central obesity, and their changes with risk of knee osteoarthritis: a nationwide population-based cohort study. Sci Rep. (2023) 13:3796. doi: 10.1038/s41598-023-30727-4, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhuo Q, Yang W, Chen J, Wang Y. Metabolic syndrome meets osteoarthritis. Nat Rev Rheumatol. (2012) 8:729–37. doi: 10.1038/nrrheum.2012.135, [DOI] [PubMed] [Google Scholar]
- 24.Muthuri S, McWilliams D, Doherty M, Zhang W. History of knee injuries and knee osteoarthritis: a meta-analysis of observational studies. Osteoarthr Cartil. (2011) 19:1286–93. doi: 10.1016/j.joca.2011.07.015, [DOI] [PubMed] [Google Scholar]
- 25.Whittaker JL, Losciale JM, Juhl CB, Thorlund JB, Lundberg M, Truong LK, et al. Risk factors for knee osteoarthritis after traumatic knee injury: a systematic review and meta-analysis of randomised controlled trials and cohort studies for the OPTIKNEE consensus. Br J Sports Med. (2022) 56:1406–21. doi: 10.1136/bjsports-2022-105496, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wang X, Perry TA, Arden N, Chen L, Parsons CM, Cooper C, et al. Occupational risk in knee osteoarthritis: a systematic review and meta-analysis of observational studies. Arthritis Care Res. (2020) 72:1213–23. doi: 10.1002/acr.24333, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Panigrahi A, Sahu P, Mohanty SS, Dandsena RS, Sahani JI, Pati S. Sleep disturbances and sleep quality among individuals diagnosed with osteoarthritis: a systematic review and meta-analysis. Front Med. (2025) 12:1653047. doi: 10.3389/fmed.2025.1653047, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mathieu S, Rat A-C, Sellam J. Prevalence and impact of sleep disorders in osteoarthritis: a systematic review and meta-analysis. Osteoarthr Cartil Open. (2025) 8:100738. doi: 10.1016/j.ocarto.2025.100738, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhou S, Wu L, Si H, Li M, Liu Y, Shen B. Association between nighttime sleep duration and quality with knee osteoarthritis in middle-aged and older Chinese: a longitudinal cohort study. Arch Gerontol Geriatr. (2024) 118:105284. doi: 10.1016/j.archger.2023.105284, [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 raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.