This study employed the Global Burden of Diseases (GBD) 2021 dataset, developed by the Institute for Health Metrics and Evaluation (IHME) at the University of Washington. The GBD study systematically evaluates 371 diseases and 88 risk factors across 204 countries and territories (1990–2021) using standardized age- and sex-adjusted methodologies. Data were obtained from the publicly accessible Global Health Data Exchange (GHDx).¹ From GBD 2021 dataset, we extracted age-standardized rates (ASRs) and absolute counts of incidence, prevalence, and disability-adjusted life years (DALYs) for OA, including site-specific subtypes (hip, knee, hand, and other OA), from 1990 to 2021. Additionally, we obtained age-specific population estimates (1990–2021) and projections (2022–2030) for China, the US, and globally, as well as the percentage of OA-related DALYs attributable to BMI (1990–2021) for overall OA, hip OA, and knee OA in these regions.

In the GBD 2021 study, OA was categorized using International Classification of Diseases, 10th Edition code B.11.2, encompassing hand, hip, knee, and other subtypes. Cases of OA are identified based on clinical symptoms and radiologic confirmation using Kellgren–Lawrence (K–L) grading system levels II–IV (19, 20). DALYs, a key epidemiological metric, measure the burden of disease in terms of the total time lived with disability by both survivors and those who have died. The metric combines premature mortality (YLLs) and disability duration (YLDs), calculated as their sum (DALYs = YLLs + YLDs) (21). Since deaths directly attributed to OA were not reported in GBD 2021, YLLs are assumed to be zero, making DALYs equivalent to YLDs. Consequently, this study utilizes DALYs to evaluate the impact of OA. High BMI (defined as >20–25 kg/m² in adults ≥20 years) is a GBD-validated risk factor for OA, with biomechanical stress explaining its stronger association with weight-bearing joints such as the knee and hip (9).

Using GBD 2021 data, this study assessed the OA burden in China, the US, and globally through incidence, prevalence, DALYs, age-standardized incidence rate (ASIR), prevalence rate (ASPR), and disability-adjusted life years (AS-DALYs). All estimates include age-standardized and mean values with 95% uncertainty intervals (UIs) (8). The Joinpoint regression model is an epidemiological tool used for trend analysis. It begins with a minimal number of joinpoint, testing their statistical significance and connecting those with similar trends into a smooth line (22). This model is applicable for analyzing temporal trends in ASIR, ASPR, and AS-DALYs of diseases (23). This study calculated annual percentage change and average annual percentage changes (AAPC) for ASPR, ASIR, and AS-DALY of OA in China, the US, and globally. Annual percentage change reflects changes over specific periods, while AAPC summarizes overall trends. An annual percentage change > 0 indicates an increase, < 0 indicates a decrease, and a value equal to 0 suggests a stable or non-significant change over time.

The APC model is commonly applied in epidemiological research, especially for analyzing chronic non-communicable diseases and cancer trends (24). This model disentangles the effects of age, period, and birth cohort in epidemiological data, enabling a more precise estimation of their independent impacts on disease incidence or mortality. It also provides insights into long-term trends in disease progression over time. This study employed APC analysis to assess OA incidence trends in China, the US, and globally, using 5-year aggregated data (1992–2021) stratified by age. The APC Web Tool was utilized to construct models and derive age-specific rate ratios (RRs) for different periods and cohorts (25).

According to GBD 2021, elevated BMI (≥25 kg/m²)—the sole identified risk factor for OA, hip OA, and knee OA—was analyzed due to its causal association with OA risk variation, with corresponding AS-DALYs reported. BAPC analysis was applied to predict OA disease burden trends from 2022 to 2030, using ASIR, ASPR, and AS-DALYs population data, and projected population figures. Statistical analyses and visualizations were performed with R software (v4.4.2) and IHME’s interactive data visualization tools, adopting a significant threshold of p < 0.05.

According to the analysis of GBD 2021 data, the burden of OA among individuals aged 55 years and older shows considerable variation across countries (Figure 1; Supplementary Table S1). Figure 1 presents the global distribution of OA incidence, prevalence, and DALYs rates across world regions, generated using IHME’s interactive data visualization tools. Among all countries, the highest levels were consistently observed in the US, Japan, and South Korea. Specifically, in 2021, the US had an incidence rate of 1973.19 per 100,000, a prevalence rate of 39030.76 per 100,000, and a DALYs rate of 1416.48 per 100,000, whereas China showed comparatively lower rates of 1507.57, 29525.33, and 1033.44 per 100,000, respectively. Other G20 countries, including India, Brazil, and South Africa, generally exhibited lower rates across these indicators.

From 1990 to 2021, the ASIR, ASPR, and AS-DALYs of OA increased in both China and the US among individuals aged 55 and older; however, China showed a faster growth rate across all indicators, as shown in Table 1. In China, ASIR of OA increased from 1311.06 (95% UI: 983.94, 1696.27) to 1497.13 (95% UI: 1239.56, 1920.49) per 100,000 population from 1990 to 2021, reflecting an AAPC of 0.46% (95% CI: 0.41, 0.50%). This contrasts with the US, where the ASIR showed a more modest increase from 1878.80 (95% UI: 1432.06, 2386.38) to 2012.44 (95% UI: 1545.37, 2546.84) per 100,000 population over the same period, with an AAPC of 0.20% (95% CI: 0.15, 0.25%). This suggests a faster increase in new OA cases in China compared to the US. Similarly, China experienced a substantial increase in the ASPR, rising from 26227.10 (95% UI: 22733.70, 29936.11) to 30035.65 (95% UI: 26182.66, 34173.46) per 100,000 population with an AAPC of 0.46% (95% CI: 0.44, 0.48%). In contrast, the ASPR in the US showed a smaller increase from 36360.90 (95% UI, 32012.11, 40895.85) to 38266.55 (95% UI, 33761.37, 42953.64) per 100,000 population (AAPC = 0.16%; 95% CI: 0.13, 0.19%). This indicates a higher overall prevalence rate of OA in the US, but a faster growth in prevalence in China. The trend in AS-DALYs followed a similar pattern. China experienced a larger increase in AS-DALYs, from 899.55 (95% UI, 430.98, 1813.47) to 1050.35 (95% UI, 502.60, 2117.86) per 100,000 population, with an AAPC of 0.52% (95% CI, 0.50, 0.54%). In comparison, AS-DALYs in the US increased from 1322.63 (95% UI, 646.60, 2667.35) to 1387.79 (95% UI, 681.85, 2793.00) per 100,000 population (AAPC = 0.15%; 95% CI: 0.12, 0.18%). In addition to age-standardized rates, the absolute number of OA cases has grown substantially, particularly in China. According to Table 2, the number of incident cases in China rose dramatically from 1.90 million in 1990 to 5.71 million in 2021, and prevalent cases expanded from 36.1 million to 111.9 million. In comparison, the US experienced a more modest increase from 0.95 million to 1.98 million incident cases, and from 19.7 million to 39.1 million prevalent cases. These findings highlight distinct epidemiological patterns: China’s massive population generates larger absolute OA burden, though with currently lower per-capita rates than the US. The accelerating growth in China’s case numbers and age-standardized rates signals an emerging public health challenge, while the US maintains persistently high prevalence. These disparities may stem from fundamental differences in population demographics, risk factor profiles (particularly obesity and aging patterns), and healthcare systems between the two countries.

In China, the AAPCs for hand and hip OA ASPR from 1990 to 2021 are 1.6% (95% CI: 1.53–1.66) and 0.88% (95% CI: 0.86–0.90), respectively (Table 1). These values are higher than the AAPCs for overall OA (0.46, 95% CI: 0.44–0.48) and knee OA (0.24, 95% CI: 0.22–0.27). This indicates a sharper increase in the burden of hand and hip OA compared to other OA subtypes in China. For ASIR, hand OA and hip OA in China also show a relatively higher AAPC. Hand OA has an AAPC of 1.22% (95% CI: 1.18–1.25), and hip OA stands at 0.91% (95% CI: 0.89–0.93), surpassing the overall OA AAPC of 0.46% (95% CI: 0.41–0.50). This further highlights the rapid growth of new cases in these specific OA subtypes. In terms of AS-DALYs, the AAPC for hand OA is 1.59% (95% CI: 1.53–1.65), and for hip OA, it is 0.87% (95% CI: 0.85–0.89). These rates are much higher compared to the AAPC for knee OA (0.23, 95% CI: 0.21–0.25) and overall OA (0.52, 95% CI: 0.50–0.54). This indicates a faster-growing disability burden from hand and hip OA among Chinese adults aged ≥55 years. In contrast, the US shows lower AAPCs across all subtypes of OA. For instance, the AAPCs for hand OA ASPR is 0.2% (95% CI: 0.14–0.26), and for hip OA, it is 0.35% (95% CI: 0.20–0.46). Similarly, ASIR and AS-DALYs for hand and hip OA have AAPCs below 0.4%, much lower than China’s values. The AAPCs for hand and hip OA in China are remarkably higher than those for other OA subtypes in the country and significantly exceed the corresponding rates in the US. These findings indicate a need for targeted public health interventions in China, focusing particularly on the rapidly increasing burden of hand and hip OA.

Regarding the ASIR of OA in individuals aged 55 and older, China exhibited an overall upward trajectory, characterized by a sharp rise between 2000 and 2005, followed by a gradual decline until 2015 and a subsequent slow rising (Figure 2). In the US, the ASIR rose rapidly from 1990 to 1995, peaked in 1995, and then slowly declined, reaching a trough in 2005 before rising again gradually (Figure 2). In China, the US, and globally, the trends in ASPR and AS-DALYs mirror those observed for ASIR. Furthermore, in both countries, all three indicators for OA are generally higher in females than in males.

As shown in Figure 3, together with Supplementary Figures S1, S2, the age-specific distribution of OA burden among individuals aged 55 years and older was analyzed by both number and rate of DALYs, incidence and prevalence across different joint sites (hip, knee, hand, and other OA) in China, the US, and globally. In China, the largest DALYs number for knee OA was observed in the 55–59 years age group. However, the DALYs rate for knee OA peaked in the 85–89 years age group, indicating a shift in the age-related impact from absolute counts to relative rates. For hand OA in China, the largest DALYs number was observed in the 65–69 years group, while its rate continued to increase with advancing age, becoming especially prominent in the oldest age groups. In the US, the highest number of DALYs from knee OA occurred in the 65–69 years group, while the DALYs rate peaked in the 75–79 years group. In the US, the proportion of hand OA-related DALYs rate was consistently higher than in China across all age groups. In both countries, hand OA-related DALYs rate increased steadily with age, and the absolute number of DALYs peaked around 65–69 years. Hip OA in both China and the US contributed a smaller proportion of total DALYs, but its rate gradually increased with age, reflecting the cumulative burden in the oldest individuals. Similarly, across all regions, the rate of DALYs due to “other OA” types showed a consistent rise with age, becoming a more significant contributor among those aged 90 years and older. Overall, knee OA remains the leading contributor to OA-related DALYs in younger older adults, while hand OA and other OA become increasingly important in the oldest age groups. These findings highlight both joint-specific and age-specific patterns of OA burden, with noticeable differences between China and the US.

Longitudinal age curves primarily describe the age effect, illustrating the trend of disease incidence as age increases (Figure 4). The analysis of age effects reveals that globally and, in the US, the incidence of OA gradually decreases with increasing age. However, in China, there is a significant peak in incidence between the ages of 75 and 85. To illustrate the period effect on OA incidence from 1992 to 2021, we present a comparative analysis across three regions: China, the US, and globally (Figure 4). The period from 2007 to 2012 serves as the reference, with an incidence rate ratio of 1. Distinct patterns emerge across the three regions. In China, incidence gradually increased from 1992, peaking between 2002 and 2007, followed by a decline and a subsequent rebound starting in the 2012–2017 period, which continues to rise. Conversely, the US exhibits a decreasing trend from 1990, reaching its lowest point between 2002 and 2007 before rising again. Globally, a steady increase in incidence is observed. The impact of cohort effects on OA incidence from 1992 to 2021 is analyzed across three regions: China, the US, and globally (Figure 4). This ratio measures the relative risk of developing OA for a specific birth cohort compared to a reference cohort, revealing the long-term influence of birth cohort on incidence. The cohort of individuals born from 1930 to 1935 serves as the reference in this analysis. A general upward trend in OA incidence is observed across birth cohorts from 1905 to 1965 in China, the US, and globally. However, a distinct pattern emerges in the US, where a peak incidence is observed in the cohort born in the 1930s, followed by a slight decline and a subsequent gradual increase from the 1940s onwards.

We analyzed the contribution of BMI to the AS-DALYs associated with hip and knee OA in individuals aged 55 and older (Table 3). For hip OA, China exhibited a substantial rise in AS-DALYs, escalating from 4.47 (−0.34, 12.50) per 100,000 population in 1990 to 10.39 (−0.86, 28.38) in 2021, accompanied by an AAPC of 2.76 (2.74, 2.79). This marked growth in disability burden was also observed in the US, albeit less pronounced, with AS-DALYs increasing from 46.41 (−4.56, 127.95) in 1990 to 60.56 (−6.59, 155.96) in 2021 and an AAPC of 0.78 (0.64, 0.88). Globally, the trend mirrored that of the US, with AS-DALYs rising from 16.78 (−1.48, 45.45) in 1990 to 20.99 (−1.98, 55.41) in 2021 and an AAPC of 0.73 (0.7, 0.75). For knee OA, China presented that AS-DALYs increased substantially from 105.50 (−8.48, 314.53) in 1990 to 202.36 (−17.87, 580.46) in 2021, with an AAPC of 2.13 (2.11, 2.15). In the US, the values rose from 261.06 (−25.32, 706.24) in 1990 to 316.51 (−34.78, 837.30) in 2021, with an AAPC of 0.68 (0.63, 0.73), while globally, they climbed from 142.68 (−12.32, 405.07) in 1990 to 197.16 (−18.42, 549.13) in 2021, with an AAPC of 1.05 (1.05, 1.06). While the US exhibits higher AS-DALYs values for hip and knee OA, China demonstrates a more rapid growth rate in AS-DALYs, indicating that the contribution of BMI to the escalating burden of OA is increasing faster in China compared to the US. This observation suggests that while BMI currently plays a more prominent role in the OA burden in the US, its impact is accelerating more rapidly in China.

To project the burden of OA in China, the US, and globally, from 2022 to 2030, we employed the BAPC prediction model to estimate ASIR, ASPR, and AS-DALYs in individuals aged 55 and older. Figure 5 presents the model’s projections, with the left y-axis showing the number of cases and the right y-axis displays the ASR data. The projections indicate a persistent rise in OA burden across China, the US, and globally through 2030. While China currently exhibits lower ASIR, ASPR, and AS-DALYs than the US, it demonstrates a higher rate of increase. Taking ASIR as an example, projections indicate that China will experience a substantial increase, from 1,505 in 2022 to 1,614 in 2030 per 100,000 population, representing a growth rate of about 7.2%. In contrast, the US is projected to see a more modest rise from 2048 to 2,143 per 100,000 population, with a growth rate of approximately 4.6%. Globally, ASIR is expected to rise from 1,612 to 1,661 per 100,000 population, reflecting a growth rate of about 3.0%. Projections indicate a persistent increase in the OA burden among individuals aged 55 and older in China, the US, and globally through 2030. While China currently exhibits lower ASIR, ASPR, and AS-DALYs than the US, it demonstrates a higher rate of increase. This faster growth in China suggests that public health interventions and healthcare resources may need to be scaled up more rapidly in China compared to the US to address the future burden of OA effectively.