The prevalence of symptomatic knee osteoarthritis in China: Results from China Health and Retirement Longitudinal Study

Purpose: To describe the prevalence of symptomatic knee osteoarthritis (SxOA) in Chinese population. Methods: China Health and Retirement Longitudinal Study (CHARLS) is a nationwide population-based longitudinal survey among Chinese retired population. Individuals (age ≥45 years) and their spouses were interviewed in 150 randomly selected counties in China during 2011-2012, with a total of 17708 subjects recruited. Trained health professionals went door to door to administer the survey questionnaires, including socio-demographic factors (i.e., age, sex, living address, education), socio-economic factors (i.e., local Gross Domestic Product per capita), and medical history. Participants were asked whether they had knee pain at the time of the interview and whether they had been diagnosed arthritis by a physician. In this analysis we defined a subject as having SxOA if he/she responded to both questions positively We divided subject’s age into 4 groups: <50, 50-59, 60-69 and ≥70 years old, and their living localities into 6 regions: East, North, North-East, North-West, South-Central, and South-West. Individuals living in Hainan, Ningxia, Taiwan, and Tibet were not selected in this survey. Education was divided into 4 categories: no formal education, elementary school only, middle/high school only, and college or more education. Subject’s economic status was divided into 3 categories: low (urban: 5530 to 20053 yuan; rural: 5530 to 16644 yuan), middle (urban: 20254 to 35767 yuan; rural: 16839 to 27969 yuan), and high (urban: 36552 to 94297 yuan; rural: 28464 to 93043 yuan). We used the PROC SURVEYFREQ procedure (SAS 9.2; SAS Institute, Cary, NC, USA) to calculate the overall and sex-specific prevalence of knee SxOA according to strata of each risk factor among all the participants. We used the PROC SURVEYLOGISTIC procedure (SAS 9.2) to examine the association of each risk factor with the prevalence of knee SxOA adjusting for other potential confounders (gender, age, education, GDP per capita, urban/rural residents and regions). Both procedures took into account the complex survey design and non-response rate of CHARLS survey in both estimates and the corresponding standard errors (S.E). Results: Included in this analysis were 17128 (96.7%) participants (mean age: 59.8 years, women: 51.6%) who had completed data on knee SxOA. As shown in Table 1, prevalence of knee SxOA was higher in women than that in men, with adjusted odds ratio (OR) = 1.76, 95% confidence interval (CI): 1.52-2.03. Knee SxOA prevalence increased with age until 70 years old then levelled off (P for trend < 0.01). Subjects who received middle/high school or college education had lower prevalence of knee SxOA than those who received no formal or only elementary school education (P<0.01). Participants who lived in the rural areas had higher prevalence of knee SxOA than those lived in the urban areas (OR=1.63, 95% CI: 1.28-2.09). Participants living in more developed areas had much lower prevalence of knee SxOA (5.1%) than those living in less developed areas (11.7%) (P<0.01). As shown in Figure 1 and Table 1, there was a difference in prevalence of knee SxOA according to geographic location (P<0.01): participants in the North and East regions had the lowest prevalence of knee SxOA, followed by those in the North-East (7.0%), South-Central (7.8%) and the North-West (10.8%) regions. Participants in the South-West region had the highest prevalence of SxOA (13.7%). Conclusions: Prevalence of knee SxOA was higher in women than that in men and increased with age until 70 years old. Subjects living in rural areas, receiving less education, and having low income had higher prevalence of knee SxOA than their counterparts. There was a wide geographical variation in prevalence of knee SxOA, with participants in the west regions having much higher prevalence of knee SxOA than those living in east regions of China. Future studies are needed to identify the potential risk factors for such regional variation in prevalence of knee SxOA.Tabled 1Table 1. Prevalence of symptomatic knee OA and adjusted ORMen(%)Women(%)Total(%)Multivariable adjusted OR (95% CI)Total5.710.38.1–Age (year)<503.36.85.21.0 (reference)50–595.79.17.41.38(1.13–1.69)60–697.514.711.11.90(1.52–2.39)≥ 705.911.58.81.42(1.12–1.81)P for trend–––<0.01EducationNo formal education9.912.211.71.0 (reference)Elementary school7.111.69.20.96(0.80–1.14)Middle/High school3.56.54.70.57(0.47–0.69)College and over0.81.61.10.15(0.07–0.35)P for trend–––<0.01GDP per capitaLow8.814.311.71.0 (reference)Middle6.411.18.80.91(0.72–1.15)High3.17.05.10.61(0.46–0.79)P for trend–––<0.01LocalityUrban2.97.25.21.0 (reference)Rural8.313.510.91.63(1.28–2.09)P value–––<0.01RegionsNorth3.87.35.41.0 (reference)East4.07.05.50.88(0.64–1.21)North-East4.29.87.01.24(0.80–1.94)South-Central5.310.37.81.21(0.87–1.68)North-West8,113.510.81.62(1.06–2.49)South-West10.117.413.71.83(1.32–2.56)P value---<0.01 Open table in a new tab