Global and Chinese burden of three major non-communicable diseases attributable to high BMI: trends from 1990 to 2021 and projections to 2050 Authors: Hao Qiao, Jia Li Published: 2026 ABSTRACT BackgroundChronic non‑communicable diseases (NCDs) are the leading cause of mortality worldwide. High body mass index (BMI) is a major modifiable risk factor for NCDs. This study aimed to compare temporal trends and age- and sex-specific patterns of high BMI-attributable NCD burden in China and globally from 1990 to 2021, disentangle the underlying drivers, and project the burden to 2050.MethodsWe used data from the Global Burden of Disease Study 2021 (GBD 2021) to estimate deaths and disability‑adjusted life years (DALYs) attributable to high BMI for neoplasms, cardiovascular diseases (CVDs), and diabetes mellitus in China and globally from 1990 to 2021. We applied Joinpoint regression, Age‑Period‑Cohort (APC) analysis, and decomposition analysis to identify temporal trends, inflection points, and driving factors, and used autoregressive integrated moving average (ARIMA) time series models to project future trends through 2050.ResultsIn 2021, age‑standardized mortality rates (ASMR) per 100,000 population in China attributable to high BMI were 2.81 (95% UI 1.20–4.76) for neoplasms, 18.80 (10.07–29.43) for CVDs, and 3.57 (1.37–5.72) for diabetes; corresponding age‑standardized DALY rates (ASDR) were 79.17 (33.82–134.14), 396.09 (197.27–638.54), and 292.46 (136.21–439.33), all higher than global averages. From 1990 to 2021, China showed positive AAPCs for neoplasms (ASMR, 2.4%; ASDR, 2.3%), CVDs (ASMR, 1.0%; ASDR, 1.1%), and diabetes (ASMR, 1.1%; ASDR, 2.3%). Compared with the corresponding global estimates, the increases were more pronounced for neoplasms, CVDs, and diabetes ASDR, whereas the diabetes ASMR trends were similar. Decomposition revealed that epidemiological change was the dominant driver in China, whereas population growth was the main driver globally, with smaller, outcome-specific contributions from aging. ARIMA projections suggested that the burden of high BMI-related NCDs in China may continue to increase through 2050, particularly for neoplasms and CVDs.ConclusionThe deaths and DALYs burden of NCDs attributable to high BMI in China and globally showed an increasing trend since 1990 and will continue increasing through 2050. Epidemiological changes were the predominant driver in China, whereas population growth was the main driver globally. Targeted screening and proactive weight-management strategies for high-risk populations may help mitigate the growing burden of high BMI-related NCDs.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s41182-026-01071-w. Introduction Chronic non-communicable diseases (NCDs) include cardiovascular diseases, cancer, and diabetes. They are major global health challenges and account for 74% of total deaths worldwide [1]. The 2021 Global Burden of Disease (GBD) study shows that NCDs caused 1.73 billion disability-adjusted life years (DALYs) globally in 2021, a 17.6% increase since 2000 [2]. This trend brings heavy pressure to health systems and socioeconomic development. The burden is growing fastest in developing countries, especially in China [3, 4]. In 2021, NCD deaths in China reached 10.451 million, accounting for 91% of all deaths and ranking first globally [4, 5]. Neoplasms, cardiovascular diseases, and diabetes mellitus are three major NCDs that contribute substantially to the global disease burden, with marked increases in their incidence and prevalence documented since 1990 [6]. However, comprehensive evidence comparing the long-term trends and drivers of the high BMI-attributable burden of these three major NCDs in China with global patterns remains limited. High body mass index (BMI) is a well‑recognized modifiable risk factor for multiple NCDs [7–11]. GBD 2021 data indicate that overweight and obesity led to 3.71 million deaths and 129 million DALYs globally in 2021. The age‑standardized DALY rate has risen by more than 15% in two decades [12, 13]. Over the past 40 years, rapid economic growth, urbanization, dietary changes, and reduced physical activity have greatly increased overweight and obesity rates in China [14, 15]. China had the largest adult population with overweight or obesity in 2021, with 402 million people [16]. Evidence shows that the global burden of NCDs attributable to high BMI has kept increasing [17–19]. These findings highlight the need to strengthen obesity‑related NCD prevention and control. Future projections suggest a continuous rise in obesity‑related health threats. Global obesity‑related cancer cases may reach 2 million by 2070 [20]. Understanding current burden and future trends is essential for effective policymaking. Several recent studies using GBD 2021 data have characterized the burden of NCDs attributable to high BMI in China, but many have focused on individual diseases or specific analytical dimensions [3, 4]. Few studies have simultaneously compared China with global patterns across neoplasms, cardiovascular diseases, and diabetes mellitus while integrating Joinpoint regression, age-period-cohort analysis, decomposition analysis, and ARIMA forecasting. To address this gap, we applied these complementary methods within a unified framework and additionally examined age- and sex-specific patterns, thereby providing an integrated assessment of temporal changes, underlying drivers, population heterogeneity, and projected NCD burden through 2050. Methods Data resource This study utilized data from the Global Burden of Disease Study 2021 (GBD 2021) [2], compiled by the Institute for Health Metrics and Evaluation (IHME). The GBD 2021 comprehensively quantifies the burden of 371 diseases and injuries and 88 associated risk factors across 204 countries and territories, spanning 21 regions and 7 super-regions from 1990 to 2021. The GBD Results Tool was utilized to gather data on annual numbers and age-standardized rates (ASRs) of mortality and DALYs of NCDs (neoplasms, cardiovascular diseases, and diabetes) caused by high BMI in China and globally from 1990 to 2021. ASRs were expressed per 100,000 population and standardized using the GBD standard population. The definition of high BMI for adults aged 20 and above is a BMI greater than 25 kg/m2. Identification of NCDs in GBD 2021 was according to the standardized diagnostic codes ICD-10 and ICD-9. Joinpoint regression analysis Joinpoint regression was used to identify time points at which the direction or magnitude of temporal trends changed significantly [21]. The method estimates the annual percentage change for each trend segment and summarizes the overall trend using the average annual percentage change (AAPC). It was selected because it can detect inflection points that a single linear trend may obscure. Analyses were conducted using the National Cancer Institute Joinpoint Regression Program (version 5.2.0), with a grid-search method and a maximum of six joinpoints. Model selection was performed using the Monte Carlo permutation test with 4499 permutations and an overall significance level of 0.05. Parametric confidence intervals were calculated for the annual percentage change and AAPC estimates, and the AAPC was estimated over the entire 1990–2021 period [21, 22]. Age-period-cohort (APC) analysis and projection An APC analysis utilizes the Poisson distribution to capture temporal trends in mortality or DALY rates by age, period, and cohort, and is extensively applied in sociological and epidemiological studies [23]. The APC model was utilized to characterize age-, period-, and cohort-related patterns in the burden of NCDs linked to high BMI. The age effects reflect the biological changes related to aging in 16 five-year age strata (20–24, 25–29, 30–34, …, 95–99), period effects capture temporal variations in six sequential 5-year intervals (1992–1996 to 2017–2021), and cohort effects denote health disparities in 21 birth cohorts (1897, 1902, 1907, …, 1997) due to differential risk exposures. Analyses were performed using the NCI APC Web Tool (https://analysistools.cancer.gov/apc/), which applies an estimable-functions approach to address the APC identifiability problem. The default central reference groups generated by the tool were used: 55–59 years for age, 2002–2006 for period, and 1947 for birth cohort. Age, period, and cohort effects were evaluated relative to their corresponding reference groups, and relative risks (RRs) were reported where applicable. Net drift represents the overall annual percentage change in the expected age-standardized rate over the study period, whereas local drift represents the annual percentage change in the age-specific rate for each age group. Positive drift values indicate increasing rates, whereas negative values indicate decreasing rates. Decomposition analysis Decomposition analysis was used to identify the drivers of changes in deaths and DALYs in China and globally between 1990 and 2021 [24]. Following the Das Gupta decomposition framework, the total change was partitioned into contributions from population growth, population aging, and epidemiological change by averaging the effects across all possible factor-substitution sequences [25]. Population aging reflected changes in age structure, population growth reflected changes in total population size, and epidemiological change represented changes in BMI-attributable disease rates independent of demographic changes. The autoregressive integrated moving average model (ARIMA Model) We projected the high BMI-attributable NCDs burden in China and globally to 2050 using the ARIMA model. The ARIMA model serves as a statistical approach for predicting and analyzing time series data [26]. The ARIMA model uses specific mathematical models to describe the characteristics of a time series based on autocorrelation analysis, and further predicts future data based on the model. The model integrates three core components: the autoregressive (AR) term correlating current values with prior observations, integration (I) converting non-stationary series to stationary, and moving average (MA) modeling errors as linear combinations of past residuals. Stationarity was assessed using the KPSS test, and ACF and PACF plots were examined. The optimal model for each series was selected using auto.arima() based on the minimum AICc. Model performance was evaluated using AIC, AICc, BIC, residual ACF plots, and the Ljung–Box test. Five-year holdout validation was conducted using 2017–2021 data, with RMSE, MAE, and MAPE as accuracy measures. The final models were fitted to the complete 1990–2021 series to forecast 2022–2050 with 95% prediction intervals. ARIMA results were analyzed and visualized using the “forecast” and “ggplot2” packages in R version 4.0.2. Statistical analysis All statistical analyses and visualizations, except Joinpoint regression and age-period-cohort analysis, were performed using R version 4.0.2. A two-sided P value <0.05 was considered statistically significant. Burden estimates are reported with 95% uncertainty intervals (UIs) directly extracted from the GBD Results Tool and were not recalculated. These intervals reflect uncertainty propagated through the GBD estimation process. The model-derived estimates from Joinpoint regression and age-period-cohort analysis are reported with 95% confidence intervals (CIs). Results Temporal trends in the burden of NCDs attributable to high BMI in China and globally from 1990 to 2021 Table 1 details deaths, DALYs, ASMR, and ASDR for high BMI-attributable neoplasms, cardiovascular diseases (CVDs), and diabetes mellitus (DM) in China from 1990 to 2021. In 2021, high BMI contributed to 58,745 (24,601–99,889) neoplasm deaths, 354,345 (188,060–564,964) CVD deaths, and 73,758 (28,819–116,526) diabetes deaths in China. The corresponding ASMR was 2.81 (1.2, 4.76), 18.8 (10.07, 29.43), and 3.57 (1.37, 5.72) per 100,000 people. Similarly, high BMI resulted in 1,658,721 neoplasm DALYs, 8,000,896 CVD DALYs, and 5,898,043 DM DALYs, with ASDR of 79.17, 396.09, and 292.46 per 100,000 people (Table 1). Table 1Burden of NCDs attributable to high BMI in China and globally, 1990 and 2021Metric/measure1990 values (95% UI)2021 values (95% UI)Percentage change in ASR, 1990–2021 (%)AAPCChinaGlobalChinaGlobalChinaGlobalChinaGlobalNeoplasms Mortality number11,556 (6,040, 18,285)137,353 (57,450, 225,297)58,745 (24,601, 99,889)356,738 (146,116, 581,012) ASMR, per 100,0001.36 (0.69, 2.15)3.66 (1.51, 6.03)2.81 (1.20, 4.76)4.18 (1.71, 6.80)106.6114.212.40 (2.20 to 2.60)0.40 (0.40 to 0.50) DALYs377,746 (203,505, 588,229)3,549,049 (1,548,429, 5,731,481)1,658,721 (693,186, 2,831,308)8,894,525 (3,751,953, 14,385,271) ASDR, per 100,00039.14 (20.46, 61.54)87.53 (37.43, 141.83)79.17 (33.82, 134.14)102.17 (43.24, 165.02)102.2716.532.30 (2.20 to 2.40)1.90 (1.90 to 2.00)Cardiovascular diseases Mortality number88,034 (61,878, 122,804)863,066 (477,898, 1,316,551)354,345 (188,060, 564,964)1,904,239 (1,072,732, 2,864,242) ASMR, per 100,00013.73 (8.44, 19.94)24.43 (13.63, 37.30)18.80 (10.07, 29.43)22.77 (12.87, 34.24)36.92−6.791.00 (0.80 to 1.20)−0.20 (− 0.40 to 0.00) DALYs2,253,093 (1,616,542, 3,195,667)21,082,657 (11,331,044, 32,461,613)8,000,896 (3,901,647, 13,000,731)45,426,879 (23,767,187, 69,603,288) ASDR, per 100,000281.04 (196.23, 396.57)535.01 (291.22, 821.15)396.09 (197.27, 638.54)529.00 (277.28, 808.64)40.94−1.121.10 (0.90 to 1.30)0.00 (− 0.20 to 0.20)Diabetes mellitus Mortality number19,982 (8,520, 31,444)238,136 (98,915, 358,326)73,758 (28,819, 116,526)723,719 (306,923, 1,075,367) ASMR, per 100,0002.51 (1.03, 4.04)6.32 (2.55, 9.64)3.57 (1.37, 5.72)8.46 (3.55, 12.62)42.2333.861.10 (1.00 to 1.30)1.00 (0.80 to 1.10) DALYs1,387,873 (621,067, 2,124,168)10,434,885 (4,873,204, 15,314,308)5,898,043 (2,766,366, 8,874,174)39,309,214 (19,146,834, 56,417,300) ASDR, per 100,000145.22 (63.37, 226.09)251.52 (115.51, 372.35)292.46 (136.21, 439.33)452.55 (220.36, 650.49)101.3979.932.30 (2.20 to 2.40)1.90 (1.90 to 2.00)Values are estimates with 95% uncertainty intervals (UIs). Deaths and DALYs are absolute counts and are rounded to whole numbers; ASMR and ASDR are expressed per 100,000 population. Percentage change applies only to ASRs and was calculated as [(ASR in 2021 − ASR in 1990)/ASR in 1990] × 100, where ASR denotes ASMR for mortality and ASDR for DALYs. AAPC represents the average annual percentage change in ASR from 1990 to 2021 and is presented with its 95% confidence interval (CI)ASMR age-standardized mortality rate, ASDR age-standardized DALY rate, DALY disability-adjusted life year Table 1Burden of NCDs attributable to high BMI in China and globally, 1990 and 2021Metric/measure1990 values (95% UI)2021 values (95% UI)Percentage change in ASR, 1990–2021 (%)AAPCChinaGlobalChinaGlobalChinaGlobalChinaGlobalNeoplasms Mortality number11,556 (6,040, 18,285)137,353 (57,450, 225,297)58,745 (24,601, 99,889)356,738 (146,116, 581,012) ASMR, per 100,0001.36 (0.69, 2.15)3.66 (1.51, 6.03)2.81 (1.20, 4.76)4.18 (1.71, 6.80)106.6114.212.40 (2.20 to 2.60)0.40 (0.40 to 0.50) DALYs377,746 (203,505, 588,229)3,549,049 (1,548,429, 5,731,481)1,658,721 (693,186, 2,831,308)8,894,525 (3,751,953, 14,385,271) ASDR, per 100,00039.14 (20.46, 61.54)87.53 (37.43, 141.83)79.17 (33.82, 134.14)102.17 (43.24, 165.02)102.2716.532.30 (2.20 to 2.40)1.90 (1.90 to 2.00)Cardiovascular diseases Mortality number88,034 (61,878, 122,804)863,066 (477,898, 1,316,551)354,345 (188,060, 564,964)1,904,239 (1,072,732, 2,864,242) ASMR, per 100,00013.73 (8.44, 19.94)24.43 (13.63, 37.30)18.80 (10.07, 29.43)22.77 (12.87, 34.24)36.92−6.791.00 (0.80 to 1.20)−0.20 (− 0.40 to 0.00) DALYs2,253,093 (1,616,542, 3,195,667)21,082,657 (11,331,044, 32,461,613)8,000,896 (3,901,647, 13,000,731)45,426,879 (23,767,187, 69,603,288) ASDR, per 100,000281.04 (196.23, 396.57)535.01 (291.22, 821.15)396.09 (197.27, 638.54)529.00 (277.28, 808.64)40.94−1.121.10 (0.90 to 1.30)0.00 (− 0.20 to 0.20)Diabetes mellitus Mortality number19,982 (8,520, 31,444)238,136 (98,915, 358,326)73,758 (28,819, 116,526)723,719 (306,923, 1,075,367) ASMR, per 100,0002.51 (1.03, 4.04)6.32 (2.55, 9.64)3.57 (1.37, 5.72)8.46 (3.55, 12.62)42.2333.861.10 (1.00 to 1.30)1.00 (0.80 to 1.10) DALYs1,387,873 (621,067, 2,124,168)10,434,885 (4,873,204, 15,314,308)5,898,043 (2,766,366, 8,874,174)39,309,214 (19,146,834, 56,417,300) ASDR, per 100,000145.22 (63.37, 226.09)251.52 (115.51, 372.35)292.46 (136.21, 439.33)452.55 (220.36, 650.49)101.3979.932.30 (2.20 to 2.40)1.90 (1.90 to 2.00)Values are estimates with 95% uncertainty intervals (UIs). Deaths and DALYs are absolute counts and are rounded to whole numbers; ASMR and ASDR are expressed per 100,000 population. Percentage change applies only to ASRs and was calculated as [(ASR in 2021 − ASR in 1990)/ASR in 1990] × 100, where ASR denotes ASMR for mortality and ASDR for DALYs. AAPC represents the average annual percentage change in ASR from 1990 to 2021 and is presented with its 95% confidence interval (CI)ASMR age-standardized mortality rate, ASDR age-standardized DALY rate, DALY disability-adjusted life year Burden of NCDs attributable to high BMI in China and globally, 1990 and 2021 Values are estimates with 95% uncertainty intervals (UIs). Deaths and DALYs are absolute counts and are rounded to whole numbers; ASMR and ASDR are expressed per 100,000 population. Percentage change applies only to ASRs and was calculated as [(ASR in 2021 − ASR in 1990)/ASR in 1990] × 100, where ASR denotes ASMR for mortality and ASDR for DALYs. AAPC represents the average annual percentage change in ASR from 1990 to 2021 and is presented with its 95% confidence interval (CI) ASMR age-standardized mortality rate, ASDR age-standardized DALY rate, DALY disability-adjusted life year Between 1990 and 2021, deaths and DALYs increased for all three disease groups in China and globally (Table 1), but age-standardized trends differed. For neoplasms, China’s ASMR increased by an absolute 1.45 per 100,000 (relative increase, 106.61%), while ASDR increased by 40.03 per 100,000 (relative increase, 102.27%); the corresponding global relative increases were 14.21% and 16.53%, respectively. For CVDs, China’s ASMR and ASDR increased by 36.92% and 40.94%, respectively, whereas the corresponding global rates declined by 6.79% and 1.12%. For diabetes, although the absolute increases were larger globally, the relative increases were greater in China (China vs. global: ASMR, 42.23% vs. 33.86%; ASDR, 101.39% vs. 79.93%). Overall, these findings indicate a substantial and increasing burden of high BMI-related NCDs in China. A comparison of the sex distribution of NCDs burden attributable to high BMI in China and globally From 1990 to 2021, sex-specific patterns of high BMI-attributable NCD burden differed across disease groups and between China and the global population. From 1990 to 2021, the disease burden has shown different trends between genders. The burden of neoplasms in China and globally has shown significant upward trends. In China, the ASDR and ASMR of women were higher than those of men before 2011, but after that, they were lower than those of men (Fig. 1A–D). Globally, the age standardization rate for women was higher than that for men continuously (Fig. 1A–D). There are different trends in cardiovascular diseases attributable to high BMI in China and globally. The disease burden in China is gradually increasing, and the disease burden in men is always higher than that of women (Fig. 1E–H). Globally, the cardiovascular diseases burden is slowly decreasing for both men and women, with a more pronounced downward trend for women (Fig. 1E–H). The burden of diabetes attributable to high BMI shows a similar gender trend in China and globally, with a higher burden in women in the early stages, and gradually, a greater disease burden in men than women in recent years (Figs. 1I-L).Fig. 1Trends in mortality, DALYs, ASMR, and ASDR of NCDs in China and globally from 1990 to 2021. The number of neoplasms deaths (A, B) and DALYs (C, D) caused by high BMI in China and globally, as well as trends in ASRs. The number of cardiovascular diseases deaths (E, F) and DALYs (G, H) caused by high BMI in China and globally, as well as trends in ASRs. The number of diabetes deaths (I, J) and DALYs (K, L) caused by high BMI in China and the world, as well as the trend of ASRs Trends in mortality, DALYs, ASMR, and ASDR of NCDs in China and globally from 1990 to 2021. The number of neoplasms deaths (A, B) and DALYs (C, D) caused by high BMI in China and globally, as well as trends in ASRs. The number of cardiovascular diseases deaths (E, F) and DALYs (G, H) caused by high BMI in China and globally, as well as trends in ASRs. The number of diabetes deaths (I, J) and DALYs (K, L) caused by high BMI in China and the world, as well as the trend of ASRs A comparison of the age distribution of the NCDs burden attributable to high BMI in China and the globally Neoplasms deaths caused by high BMI in China and globally exhibit a similar pattern in age distribution, both increasing with age and reaching their peak in the 55–69 age group; However, there are differences in gender distribution: in China, the neoplasm mortality rate is higher in middle-aged women aged 50–74 than in men, but the disease burden is higher in men after the age of 75. And globally, the disease burden of women over 45 years old has always been higher than that of men (Fig. 2A–H). In terms of cardiovascular diseases, deaths and DALYs caused by high BMI in China and globally also increase with age, with peak deaths occurring in the 70–74 and 65–69 age groups, respectively. Moreover, the disease burden peak in males generally appears earlier than in females (Fig. 2I–P). For diabetes, deaths and DALYs increased with age in both China and globally, with deaths peaking at ages 70–74 in China and 65–69 globally (Fig. 2Q–X).Fig. 2Comparison of the age distribution of the chronic NCDs burden caused by high BMI in China and globally in 2021. Deaths (A–D) and DALYs burden (E–H) of high BMI-related neoplasms in different age and gender groups in China and globally. Deaths (I–L) and DALYs burden (M–P) of high BMI-related cardiovascular diseases in different age and gender groups in China and globally. Deaths (Q–T) and DALYs burden (U–X) of high BMI-related diabetes in different age and gender groups in China and globally Comparison of the age distribution of the chronic NCDs burden caused by high BMI in China and globally in 2021. Deaths (A–D) and DALYs burden (E–H) of high BMI-related neoplasms in different age and gender groups in China and globally. Deaths (I–L) and DALYs burden (M–P) of high BMI-related cardiovascular diseases in different age and gender groups in China and globally. Deaths (Q–T) and DALYs burden (U–X) of high BMI-related diabetes in different age and gender groups in China and globally Joinpoint regression analysis results Figure 3 shows the time trend of ASMR and ASDR of chronic diseases (neoplasms, cardiovascular diseases, diabetes) caused by high BMI in China and globally from 1990 to 2021. In terms of neoplasm, the AAPC (95% CI) of ASMR and ASDR in China was 2.40 (2.20–2.60) and 2.30 (2.20–2.40), respectively (P < 0.05). The annual percent change was positive in all stages (P < 0.05), indicating a continuous and steady increase in burden (Table S1, Fig. 3A, B). However, the global AAPC is relatively low, with ASMR and ASDR being 0.40 (0.40–0.50) and 1.90 (1.90–2.00), respectively (P < 0.05) (Table 1, Fig. 3A, B). Overall, the burden of neoplasms in China and globally is on the rise, but the increase in China is more significant, especially during the periods of 1997–2001 and 2014–2021.Fig. 3Joinpoint regression analysis of age-standardized rates of high BMI-attributable NCDs in China and globally, 1990–2021. Panels show the age-standardized mortality rate (ASMR) and age-standardized disability-adjusted life year rate (ASDR) for neoplasms (A, B), cardiovascular diseases (C, D), and diabetes mellitus (E, F). Blue circles and lines represent China, whereas green triangles and lines represent global estimates. Black stars and inverted triangles indicate the joinpoints identified for China and the global population, respectively. The annual percentage change for each trend segment is presented in the corresponding legend; asterisks indicate that the annual percentage change was significantly different from zero (P < 0.05). Rates are expressed per 100,000 population. BMI body mass index, ASMR age-standardized mortality rate, ASDR age-standardized DALY rate, DALY disability-adjusted life year Joinpoint regression analysis of age-standardized rates of high BMI-attributable NCDs in China and globally, 1990–2021. Panels show the age-standardized mortality rate (ASMR) and age-standardized disability-adjusted life year rate (ASDR) for neoplasms (A, B), cardiovascular diseases (C, D), and diabetes mellitus (E, F). Blue circles and lines represent China, whereas green triangles and lines represent global estimates. Black stars and inverted triangles indicate the joinpoints identified for China and the global population, respectively. The annual percentage change for each trend segment is presented in the corresponding legend; asterisks indicate that the annual percentage change was significantly different from zero (P < 0.05). Rates are expressed per 100,000 population. BMI body mass index, ASMR age-standardized mortality rate, ASDR age-standardized DALY rate, DALY disability-adjusted life year There are different patterns of cardiovascular diseases in China and globally. The AAPC (95% CI) of ASMR and ASDR in China is 1.00 (0.80–1.20) and 1.10 (0.90–1.30), respectively (Table 1). Although they briefly decreased from 1990 to 1998, they showed an overall fluctuating upward trend, with a significant increase from 2007 to 2011 (Table S1, Fig. 3C, D). In contrast, the global AAPC of ASMR and ASDR was − 0.20 (− 0.40 to 0.00) and 0.00 (− 0.20 to 0.20), respectively. The mortality rate significantly decreased from 2003 to 2021 (P < 0.05) (Table 1, Table S1, and Fig. 3C, D). In terms of diabetes, the AAPC of ASMR and ASDR in China is 1.10 (1.00–1.30) and 2.30 (2.20–2.40) (P < 0.05), respectively (Table 1). The mortality rate tends to be stable after rising, but the burden of DALYs continues to grow (Table S1, Fig. 3E, F). The global AAPCs of ASMR and ASDR is 1.00 (0.80–1.10) and 1.90 (1.90–2.00), respectively (P < 0.05) (Table 1, Fig. 3E, F). The data show that the burden of diabetes in China and the world continues to rise, and the rising trend in China is more significant. Impact of age, period, and cohort on death and DALY rates of NCDs due to high BMI We used the Age-period-cohort model to investigate the role of age, period, and cohort in death and DALYs of high BMI-related NCDs (Table S2 and Fig. 4). In China, positive net drifts were observed for neoplasms and cardiovascular diseases, with values of 2.66% and 2.73% for mortality and DALYs related to neoplasms, and 1.80% and 2.01% for cardiovascular diseases, respectively (P < 0.05). After adjustment for period and cohort effects, mortality and DALY rates for neoplasms and cardiovascular diseases generally increased with age. The period effects became consistently higher than the reference level after 2005, while the cohort effects increased among cohorts born after 1947. For diabetes, the net drifts were 1.00% for mortality and 2.67% for DALYs (P < 0.05). Age effects generally increased with age. The period effects for DALYs were elevated after 2004, and the cohort effects were higher than the reference level among cohorts born after 1947. In contrast, the cohort effects for diabetes mortality generally remained below 1 relative to the 1947 reference cohort. Detailed APC estimates are provided in Table S2.Fig. 4Age-period-cohort analysis of mortality and DALY rates attributable to high BMI-related NCDs in China. A–X Net drifts and Local drifts (Temporal Trends, gray); Longitudinal Age Curve (Age Effect, red); Period (blue) and cohort (green) effects. For each disease, the upper row presents mortality and the lower row presents DALY rates. From left to right, the panels show local drift with net drift, the longitudinal age curve, period rate ratios, and cohort rate ratios. Open circles indicate point estimates, solid lines connect the estimated values, and shaded areas represent 95% confidence intervals. In the local-drift panels, the horizontal line indicates the overall net drift. In the period and cohort panels, the horizontal reference line at RR = 1 indicates no difference from the reference period or birth cohort. The reference categories were 55–59 years for age, 2002–2006 for period, and the 1947 birth cohort (RR = 1) Age-period-cohort analysis of mortality and DALY rates attributable to high BMI-related NCDs in China. A–X Net drifts and Local drifts (Temporal Trends, gray); Longitudinal Age Curve (Age Effect, red); Period (blue) and cohort (green) effects. For each disease, the upper row presents mortality and the lower row presents DALY rates. From left to right, the panels show local drift with net drift, the longitudinal age curve, period rate ratios, and cohort rate ratios. Open circles indicate point estimates, solid lines connect the estimated values, and shaded areas represent 95% confidence intervals. In the local-drift panels, the horizontal line indicates the overall net drift. In the period and cohort panels, the horizontal reference line at RR = 1 indicates no difference from the reference period or birth cohort. The reference categories were 55–59 years for age, 2002–2006 for period, and the 1947 birth cohort (RR = 1) Decomposition analysis of changes in deaths and DALYs of high BMI-related NCDs Table 2 and Fig. 5 present the decomposition analysis results of deaths and DALYS attributed to three population-level factors: aging, growth, and epidemiologic change, both globally and in China. Population growth accounts for the major increase in global high BMI-related neoplasm deaths and DALYs (68.74%), with epidemiological changes also contributing (34.32%), and aging exerting a negative impact (− 3.16%). In China, the main driving factors for the increase in high BMI-related neoplasm deaths and DALYs are epidemiological changes (81.91%), followed by population growth (20.34%). Globally, population growth (111.71%) is the dominant factor leading to an overall increase in deaths and DALYs, and epidemiological changes have had a weak effect on reducing high BMI-related cardiovascular diseases in both deaths (− 9.21%) and DALYs (− 8.77%). In China, the increased deaths and DALYs of high BMI-related cardiovascular diseases are mainly due to the evolving disease prevalence patterns (73.29%) and population growth (34.94%), with epidemiological changes playing a more prominent role. In both the global and China, aging is an important contributor to the increase in DALYs of cardiovascular diseases related to high BMI (6.5% and 9.58%), whereas it contributed negatively to the number of deaths (− 2.50% globally and − 8.23% in China). Epidemiological changes and population expansion are contributing to the increase in deaths and DALYs from high BMI-related diabetes in China and worldwide. Aging has increased global deaths (2.65%) and DALYs (1.24%), while in China, it has shown the opposite effect (deaths: − 2.21%, DALYs: − 0.19%). Table 2Decomposition analysis of changes in deaths and DALYs of high BMI-related NCDsDiseaseLocationMeasureOverall differenceChange due to population-level determinants (% contribution to the total change)AgingPopulation growthEpidemiological changeNeoplasmsChinaDeaths1,061,157.74− 23,860.40 (− 2.25%)215,835.49 (20.34%)869,182.66 (81.91%)NeoplasmsGlobalDeaths8,293,223.87− 262,064.79 (− 3.16%)5,700,872.03 (68.74%)2,854,416.64 (34.32%)NeoplasmsChinaDALYs16,955,159.45− 215,094.53 (− 1.27%)3,521,479.29 (20.77%)13,648,774.69 (80.50%)NeoplasmsGlobalDALYs112,877,701.72− 1,848,816.86 (− 1.64%)78,567,072.04 (69.60%)36,159,446.54 (32.03%)Cardiovascular diseasesChinaDeaths12,403,373.07− 1,021,192.15 (− 8.23%)4,333,732.75 (34.94%)9,090,832.47 (73.29%)Cardiovascular diseasesGlobalDeaths46,325,917.61− 1,157,965.38 (− 2.50%)51,749,143.37 (111.71%)− 4,265,260.38 (− 9.21%)Cardiovascular diseasesChinaDALYs149,050,911.0414,274,479.65 (9.58%)42,431,408.34 (28.47%)92,345,023.05 (61.96%)Cardiovascular diseasesGlobalDALYs562,673,817.7436,570,373.05 (6.50%)575,464,022.18 (102.27%)− 49,360,577.49 (− 8.77%)Diabetes mellitusChinaDeaths1,652,200.07− 36,570.01 (− 2.21%)417,630.46 (25.28%)1,271,139.62 (76.94%)Diabetes mellitusGlobalDeaths16,024,652.77424,677.24 (2.65%)10,010,859.34 (62.47%)5,589,116.19 (34.88%)EChinaDALYs53,930,949.48− 101,001.72 (− 0.19%)12,117,507.82 (22.47%)41,914,443.39 (77.72%)Diabetes mellitusGlobalDALYs530,958,460.056,585,818.34 (1.24%)230,481,483.54 (43.41%)293,891,158.17 (55.35%)Values represent the absolute contribution of each component to the total change in deaths or DALYs, with percentage contributions shown in parentheses. Negative values indicate an offsetting contribution. Percentage contributions may exceed 100% when other components contribute negativelyDALY disability-adjusted life year, ASDR age-standardized DALY rate, DALY disability-adjusted life yearFig. 5Change in deaths and DALYs of high BMI-related NCDs decomposed by population aging, population growth, and epidemiological change from 1990 to 2021 in China and globally. A Death and B DALYs of neoplasm. C Death and D DALYs of cardiovascular diseases. E Death and F DALYs of diabetes. The black dots indicate the total value of change attributable to all three components Table 2Decomposition analysis of changes in deaths and DALYs of high BMI-related NCDsDiseaseLocationMeasureOverall differenceChange due to population-level determinants (% contribution to the total change)AgingPopulation growthEpidemiological changeNeoplasmsChinaDeaths1,061,157.74− 23,860.40 (− 2.25%)215,835.49 (20.34%)869,182.66 (81.91%)NeoplasmsGlobalDeaths8,293,223.87− 262,064.79 (− 3.16%)5,700,872.03 (68.74%)2,854,416.64 (34.32%)NeoplasmsChinaDALYs16,955,159.45− 215,094.53 (− 1.27%)3,521,479.29 (20.77%)13,648,774.69 (80.50%)NeoplasmsGlobalDALYs112,877,701.72− 1,848,816.86 (− 1.64%)78,567,072.04 (69.60%)36,159,446.54 (32.03%)Cardiovascular diseasesChinaDeaths12,403,373.07− 1,021,192.15 (− 8.23%)4,333,732.75 (34.94%)9,090,832.47 (73.29%)Cardiovascular diseasesGlobalDeaths46,325,917.61− 1,157,965.38 (− 2.50%)51,749,143.37 (111.71%)− 4,265,260.38 (− 9.21%)Cardiovascular diseasesChinaDALYs149,050,911.0414,274,479.65 (9.58%)42,431,408.34 (28.47%)92,345,023.05 (61.96%)Cardiovascular diseasesGlobalDALYs562,673,817.7436,570,373.05 (6.50%)575,464,022.18 (102.27%)− 49,360,577.49 (− 8.77%)Diabetes mellitusChinaDeaths1,652,200.07− 36,570.01 (− 2.21%)417,630.46 (25.28%)1,271,139.62 (76.94%)Diabetes mellitusGlobalDeaths16,024,652.77424,677.24 (2.65%)10,010,859.34 (62.47%)5,589,116.19 (34.88%)EChinaDALYs53,930,949.48− 101,001.72 (− 0.19%)12,117,507.82 (22.47%)41,914,443.39 (77.72%)Diabetes mellitusGlobalDALYs530,958,460.056,585,818.34 (1.24%)230,481,483.54 (43.41%)293,891,158.17 (55.35%)Values represent the absolute contribution of each component to the total change in deaths or DALYs, with percentage contributions shown in parentheses. Negative values indicate an offsetting contribution. Percentage contributions may exceed 100% when other components contribute negativelyDALY disability-adjusted life year, ASDR age-standardized DALY rate, DALY disability-adjusted life year Decomposition analysis of changes in deaths and DALYs of high BMI-related NCDs Values represent the absolute contribution of each component to the total change in deaths or DALYs, with percentage contributions shown in parentheses. Negative values indicate an offsetting contribution. Percentage contributions may exceed 100% when other components contribute negatively DALY disability-adjusted life year, ASDR age-standardized DALY rate, DALY disability-adjusted life year Change in deaths and DALYs of high BMI-related NCDs decomposed by population aging, population growth, and epidemiological change from 1990 to 2021 in China and globally. A Death and B DALYs of neoplasm. C Death and D DALYs of cardiovascular diseases. E Death and F DALYs of diabetes. The black dots indicate the total value of change attributable to all three components Predicting the ASDR and ASMR in NCDs due to high BMI The ARIMA model was used to predict the trends of NCDs due to high BMI in China and globally by 2050. All selected ARIMA models showed no significant residual autocorrelation according to the Ljung–Box tests (P = 0.240–0.994). Model selection criteria and 5-year validation results are summarized in Table S3. Overall, ASMR and ASDR of neoplasms, cardiovascular diseases, and diabetes attributable to high BMI continue to rise globally and in China, with the notable exception of global cardiovascular diseases trends, which showed no significant change (Table S4 and Fig. 6A–L). The increasing trends for neoplasms and cardiovascular diseases are particularly pronounced in China. It is expected that by 2050, the ASMR and ASDR of high BMI-related neoplasms in China will exceed global levels, reaching 4.86 and 137.09 cases per 100,000 people, respectively (Fig. 6A–D). The ASMR and ASDR attributed to cardiovascular diseases will reach 26.26 and 576.72, respectively (Fig. 6E–H). The ASMR and ASDR of high BMI-related diabetes globally will continue to exceed those in China, with an estimated 10.37 and 697.31 cases per 100,000 people (Fig. 6I–L). These predictions highlight the severe situation of high BMI-related NCDs burden in China and globally through 2050.Fig. 6Projected trends of the ASMR and ASDR of NCDs due to high BMI in China and globally over the next 29 years. The temporal trends of ASMR and ASDR of neoplasms (A–D), cardiovascular diseases (E–H), and diabetes (I–J) attributed to high BMI from 1990 to 2050 in China and globally. The blue line represents the true trend in NCDs burden from1990 to 2021; the red dashed line and shaded area represent the projected trend and its 95% prediction interval Projected trends of the ASMR and ASDR of NCDs due to high BMI in China and globally over the next 29 years. The temporal trends of ASMR and ASDR of neoplasms (A–D), cardiovascular diseases (E–H), and diabetes (I–J) attributed to high BMI from 1990 to 2050 in China and globally. The blue line represents the true trend in NCDs burden from1990 to 2021; the red dashed line and shaded area represent the projected trend and its 95% prediction interval Discussion This study systematically analyzed the burden of three major NCDs attributable to high BMI in China and globally from 1990 to 2021. The relative increases in neoplasm and cardiovascular diseases burden were generally more pronounced in China than in the corresponding global patterns, whereas the diabetes ASMR trends were similar. Joinpoint regression identified key inflection points in mortality and DALY trends. Decomposition analysis showed that the increasing burden in China was mainly associated with epidemiological change and population growth. Age-period-cohort analysis indicated that later birth cohorts generally faced higher relative risks of high BMI-related NCDs. ARIMA projections suggested that the burden in China may continue to increase through 2050. By examining temporal changes, population heterogeneity, underlying drivers, and future trajectories within a unified China and global framework, our study extends previous descriptive analyses and provides evidence to inform targeted strategies for reducing the burden attributable to high BMI in China. Our study confirms that the burden of high BMI-related neoplasms, cardiovascular diseases, and diabetes in China has increased rapidly from 1990 to 2021, with a growth rate significantly higher than the global average. China’s more pronounced increase may partly reflect three interrelated transitions accompanying rapid socioeconomic development: accelerated urbanization, a shift from agricultural and manual work toward less physically demanding occupations, and dietary westernization. Longitudinal studies have documented substantial declines in occupational physical activity and increases in the proportion of energy derived from animal-source foods, together with rising consumption of red and processed meat and sugar-sweetened beverages [27–31]. These changes may have reduced daily energy expenditure and increased obesogenic dietary exposure, thereby contributing to the rapid increase in high BMI-related burden. In contrast, many high-income countries have achieved stable or declining cardiovascular diseases burden related to high BMI through long-term prevention programs including lipid control, blood pressure management, and tobacco control [32, 33]. Decomposition analysis further revealed different drivers in China and globally. Epidemiological change was the dominant driver in China, whereas population growth played a leading role globally. The negative contribution of aging to mortality in several analyses may reflect shifts in population composition toward age groups with comparatively lower high BMI-attributable mortality rates, which partially offset the effects of population growth and epidemiological change. Meanwhile, previous studies have shown that increasing cancer risks in younger birth cohorts, together with population aging, may contribute to the growing obesity-related cancer burden [34, 35]. Taken together, these findings support a life-course approach to the prevention of high BMI-related NCDs, with particular emphasis on early intervention in younger populations. The burden of high BMI-related NCDs differs by sex and age. In China, the burden of neoplasms shifted from being higher in women to being higher in men around 2011, while a similar transition was observed for diabetes in recent years. This crossover may partly reflect the more rapid increase in BMI among Chinese men than women; nationally representative surveys showed that BMI increased faster in men from 2004 to 2018, reversing the earlier sex pattern in obesity prevalence [36]. However, this crossover was descriptive, and no formal statistical comparison between sexes was performed. Men generally carry a higher burden, especially at older ages, while women aged 50 to 79 also experience a substantial burden. Biological, behavioral, and social factors together may contribute to these differences. Men tend to accumulate more visceral fat, which promotes inflammation, insulin resistance and higher risks of diabetes, cardiovascular diseases, and cancer [37, 38]. Men also have higher rates of smoking and drinking, which may interact with obesity to elevate disease risks [39]. Estrogen helps regulate immunity, reduce inflammation, and improve insulin sensitivity, thus lowering NCDs risks in premenopausal women [40]. The decrease in estrogen after menopause may explain the peak burden among women aged 60–69. Obesity prevention policies and interventions should therefore account for sex differences. Programs targeting men may focus on smoking cessation, reduced alcohol consumption, and weight management, whereas programs for middle-aged and postmenopausal women may prioritize weight control and metabolic health management. Age-period-cohort analysis shows that younger generations in China face a faster increase in high BMI-related NCDs burden than older generations. Period effects show that relative risk has remained above 1.0 since 2005, suggesting that risk exposure in China has increased continuously in recent decades, even with the implementation of public health programs such as Healthy China 2030. This pattern reflects the growing obesity epidemic among young people, mainly driven by unhealthy diets and low physical activity [15, 41]. Early exposure to obesity-related risks also increases long-term cumulative cancer risk in younger generations [42, 43]. These findings support the importance of early obesity intervention in children and adolescents [44]. Sustained surveillance and trend projection are essential for effective disease control. Our projections to 2050 suggest a continuing heavy burden. By 2050, age-standardized mortality rate and age-standardized disability-adjusted life year rate of high BMI-related neoplasms in China will exceed global levels. China already has the largest number of overweight and obese adults in the world, and this number is expected to rise further by 2050 [16]. Previous projections suggest that the medical cost of overweight and obesity in China will increase substantially by 2030, placing heavy pressure on the health system [45]. In response, China has established a multilevel policy framework. The Healthy China 2030 Planning Outline and the National Nutrition Plan n provide overarching strategic direction by promoting healthy diets, physical activity, and weight management [46, 47]. These objectives are being translated into coordinated action through the Action Plan for the “Weight Management Year” Initiative, which promotes weight monitoring and nutrition and physical activity interventions across families, schools, workplaces, communities, and health-care institutions [48]. Clinical implementation has been further strengthened through the Notice on the Establishment and Management of Healthy Weight Management Clinics, with more than 5500 secondary- and tertiary-level medical institutions having established such clinics by February 2026 [49]. Effective implementation of these policies, with particular attention to younger cohorts and other high-risk populations, will be essential for mitigating the projected long-term burden of high BMI-related NCDs in China. This study provides global comparisons, uses robust statistical methods, and presents long-term projections based on GBD 2021 data. Several limitations should be noted. First, GBD estimates depend on underlying modeling assumptions and variations in the availability and quality of source data. Because the analyses relied on aggregated population-level data, the observed associations may be affected by ecological fallacy and cannot be interpreted as individual-level or causal relationships. Residual confounding and interactions between high BMI and other behavioral, metabolic, and socioeconomic factors may also remain unaccounted for. Second, the ARIMA projections assume that historical temporal patterns will remain stable through 2050 and do not incorporate future changes in high BMI exposure, prevention policies, diagnostic practices, treatment advances, or unforeseen events. Forecast uncertainty increases with the projection horizon, as reflected by the widening 95% prediction intervals; therefore, estimates for 2050 should be interpreted as indicative trends rather than precise predictions. In addition, because this study was based on GBD 2021, the burden estimates may change with subsequent GBD releases. Finally, the GBD definition of high BMI may differ from clinical definitions, and subnational variations within China were not examined. Future studies should incorporate regional and individual-level data and evaluate the effectiveness and economic implications of interventions. Conclusion From 1990 to 2021, the burden of NCDs attributable to high BMI in China increased dramatically, with growth rates exceeding global levels. Epidemiological change was the primary driver in China. Model-based projections suggest that the burden may continue to rise through 2050 if historical trends persist. Strengthened public health policies, early intervention, and targeted screening are critical to mitigate China’s growing obesity‑related NCD epidemic. Supplementary Information Supplementary Material 1: Table S1. Segment-specific annual percentage changes from Joinpoint regression analyses of high BMI-attributable ASMR and ASDR in China and globally, 1990–2021. Table S2. Age-period-cohort parameter estimates for high BMI-attributable mortality and DALY rates in China. Table S3. Model selection, residual diagnostics, and five-year validation of ARIMA models for high BMI-attributable ASMR and ASDR in China and globally. Table S4. Observed and ARIMA-projected age-standardized mortality and DALY rates attributable to high BMI for neoplasms, cardiovascular diseases, and diabetes mellitus in China and globally, 1990–2050. Supplementary Material 1: Table S1. Segment-specific annual percentage changes from Joinpoint regression analyses of high BMI-attributable ASMR and ASDR in China and globally, 1990–2021. Table S2. Age-period-cohort parameter estimates for high BMI-attributable mortality and DALY rates in China. Table S3. Model selection, residual diagnostics, and five-year validation of ARIMA models for high BMI-attributable ASMR and ASDR in China and globally. Table S4. Observed and ARIMA-projected age-standardized mortality and DALY rates attributable to high BMI for neoplasms, cardiovascular diseases, and diabetes mellitus in China and globally, 1990–2050.