According to the World Health Organization (WHO), one in eight people worldwide lived with obesity in 2022, with the global prevalence more than doubling between 1990 and 2022 (1). This trend demonstrates that the rising prevalence of high BMI worldwide has emerged as a critical public health challenge. In addition to the well-established associations with cardiovascular diseases, kidney diseases, and diabetes (2–4), growing clinical evidence indicates a positive correlation between high body mass index (BMI) and cancers such as gallbladder and biliary tract cancer, pancreatic cancer, and leukemia, among others (5–8). This suggests that BMI is an important factor in the onset and progression of cancer.

The impact of high-BMI-attributable diseases extends beyond impairing patients’ physical and mental health as well as daily functioning. It also exerts profound effects on families and overall social development due to rising healthcare expenditures, premature mortality, and productivity losses caused by work absenteeism (9, 10). According to WHO research findings, if current prevention and control measures remain unchanged, the global economic losses attributable to overweight and obesity are projected to escalate to USD 3 trillion annually by 2030 (1).

Several previous studies have investigated the cancer burden attributable to high BMI; however, these analyses were based on the 2019 Global Burden of Disease (GBD) database (11, 12). In the latest 2021 GBD database, the number of cancer types attributable to high BMI has changed from the previously reported 13 to 12, with esophageal cancer no longer included. This modification has consequently altered the proportional contributions among the remaining obesity-related cancers. Although there are studies that utilized the 2021 database, the first focused on decomposition analysis to assess the key contributors to the increasing cancer burden attributable to overweight (13), while the second adopted a holistic approach to examine age, sex, and SDI-regional disparities in BMI-associated cancer burden without detailed stratification by specific cancer types (14). Meanwhile, previous studies have not analyzed the proportional contribution of different cancer types across various age groups. Therefore, this study utilizes the latest 2021 GBD database to investigate the burden of different cancer types attributable to high BMI, stratified by geographic region, economic region, sex, and age group. Additionally, it projects the trends in cancer burden over the next decade. Furthermore, when studying cancers attributable to BMI, the findings may be influenced by various confounding factors, such as lifestyle variables (physical activity and smoking), which could obscure the establishment of definitive causal inferences. To address these limitations and rigorously evaluate the causal relationship between BMI and cancer risk, we employed an innovative Mendelian randomization approach to minimize confounding bias in our study, thereby strengthening the validity of our findings (15).

To the best of our knowledge, this represents the most comprehensive study to date, integrating GWAS and GBD databases to investigate both the causal relationships between BMI and various cancers and the disease burden attributable to high BMI. This study, utilizing the 2021 GBD database, found that the global cancer burden attributable to high BMI showed a notable upward trend between 1990 and 2021. This trend exhibited notable heterogeneity across geographic regions, economic regions, cancer types, sex, and age groups. We also found that the burden level was closely associated with socioeconomic development, with high SDI regions exhibiting a greater overall burden than low-SDI regions. Finally, we used the ARIMA model to predict the trends of various cancers attributable to high BMI over the next decade.

Firstly, our Mendelian randomization analysis confirmed causal associations between BMI and multiple malignancies (colon and rectal cancer, liver cancer, gallbladder and biliary tract cancer, pancreatic cancer, breast cancer, uterine cancer, ovarian cancer, kidney cancer, Lymphoma, multiple myeloma, and leukemia), while demonstrating no notable causal relationship with thyroid cancer. This finding appears inconsistent with the results reported in the GBD database. Integrating the characteristics of Mendelian randomization and the GBD database yields two key conclusions: First, the oncogenesis of thyroid cancer involves multifactorial determinants. If BMI-associated SNPs influence thyroid cancer development through non-BMI-mediated pathways (modulating inflammatory responses or hormonal regulation) (39). This pleiotropic effect may attenuate the estimated causal association between BMI and cancer risk. Secondly, while the GBD incorporates detailed thyroid cancer subtypes (papillary carcinoma, follicular carcinoma, and so on) in its statistical compilation, MR analyses typically treat thyroid cancer as a unitary endpoint. This methodological discrepancy may account for the null association observed between BMI and thyroid cancer.

In 2021, high BMI contributed to 356,738 global cancer deaths and 8,894,525 DALYs, marking a notable increase compared to 1990. Meanwhile, the continued rise in ASMR and ASDR (with EAPCs of 0.35 and 0.42, respectively) indicates that the cancer burden attributable to high BMI continues to worsen, even after accounting for population growth and aging. Obesity, as a risk factor for cancer, promotes tumorigenesis and progression through multiple mechanisms, including chronic inflammation (40, 41), insulin resistance (42), and disruption of gut microbiota (43). Over the past 32 years, pancreatic cancer and liver cancer demonstrated the most pronounced increases in ASR among the 12 high BMI-attributable cancers. This phenomenon may be attributed to the increasingly prominent role of obesity-related metabolic dysregulation in the pathogenesis of these cancers. For instance, high BMI can mediate the occurrence of pancreatic cancer through C-reactive protein (CRP) as an intermediary factor (44, 45). And fat cells provide fatty acids to tumor cells, which generate energy that supports the growth and spread of cancer cells (46). Moreover, the global epidemic of non-alcoholic fatty liver disease (NAFLD) may be a notable driver of the rising burden of liver cancer, particularly in regions where obesity coexists with hepatitis virus infection (47–49). In contrast, the ASR burden of gallbladder and biliary tract cancer and leukemia showed a decline (EAPC < 0). The decline in gallbladder and biliary tract cancer may be attributed to the widespread adoption of health education and cholecystectomy (50). In contrast, the etiology of leukemia is primarily linked to environmental or genetic factors. In recent years, with economic development, increased attention has been given to environmental protection, and high BMI has been shown to have a relatively low contribution to its incidence (51). Notably, the ASMR for colon and rectal cancer and thyroid cancer remained relatively stable. The stability in colon and rectal cancer may reflect how advancements in screening technologies have partially counterbalanced obesity-associated risks (52, 53), while thyroid cancer’s relatively indolent biological characteristics likely contribute to its stable ASMR profile (54). Understanding these differences assists in formulating targeted prevention and control policies based on the prevalence patterns of major cancers. For example, for rapidly increasing cancer types such as pancreatic and liver cancers, targeted screening should be implemented in obese populations, including liver function tests combined with ultrasound surveillance for NAFLD progression, and regular imaging examinations for high-risk individuals. Additionally, public health education initiatives should be strengthened to improve awareness of the cancer risks associated with obesity.

The cancer burden attributable to high BMI exhibits notable variations across geographic regions and levels of economic development. From a geographical perspective, South Asia recorded the steepest upward trends in mortality and disability burden from 1990 to 2021, while Eastern Europe demonstrated the slowest increase. The observed disparity may reflect regional differences in the stages of obesity epidemic development, availability of healthcare resources, and effectiveness of cancer prevention policy implementation. For example, the rapidly growing cancer burden in South Asia may be linked to nutrition transition driven by accelerated urbanization, characterized by widespread adoption of high-sugar, high-fat diets and reduced physical activity, while health education and policy interventions lag (55, 56). Across socioeconomically distinct regions, notable variations exist in both the overall cancer burden and composition of cancer burden as stratified by SDI levels. High-SDI regions predominantly bear the burden of colon and rectal cancer, whereas low-SDI regions are primarily affected by liver cancer. This phenomenon may be associated with the prevalent consumption of high-fat, low-fiber, and ultra-processed foods in high-SDI countries, which elevates colon and rectal cancer risk (57). In contrast, low-SDI regions face expanded liver cancer burdens due to higher exposure to aflatoxin-contaminated foods (58), prevalent hepatitis B virus infections (59), and compounding risk factors like obesity. These findings enable context-specific adaptation of prevention and control strategies. For example, in low SDI regions, relevant institutions can be established to promote “food structural reform,” restricting the production and sale of high-calorie foods while increasing subsidies for healthy foods. Additionally, integrated interventions combining hepatitis prevention and obesity management could be implemented. For example, integrating weight monitoring into HBV patient follow-up protocols and enhancing hepatitis B vaccination coverage among populations with high BMI could be implemented.

Notably, the persistent rise in obesity-related cancer burden reflects deeper structural barriers. In high-SDI regions, marketing strategies of the food industry (such as targeted advertising of ultra-processed foods) directly contribute to the proliferation of energy-dense dietary cultures (60). In low-SDI regions, both the accessibility and affordability of healthy foods are severely constrained. Under the combined effects of poverty, imbalanced food systems, and climate change, nutrient-dense foods (fruits and vegetables) remain relatively expensive, while energy-dense but nutrient-poor processed foods become more accessible (60). This dynamic exacerbates the double burden of malnutrition, where stunting and obesity coexist. Future interventions must directly address these challenges. Policy-level interventions: Enact legislation to restrict the marketing of unhealthy food. Economic leverage: Implement fruit and vegetable subsidies in low-SDI regions.

Our analysis revealed that in 2021, women bore approximately twice the cancer-related mortality and DALY burden compared to men. The predominant cancer burden among women stems from breast and uterine cancers, likely attributable to female-specific physiological characteristics and hormonal profiles. The aromatase in adipose tissue converts androgens into estrogens, which is closely related to the development and progression of hormone-dependent breast cancer (estrogen and progesterone receptor positive) (61, 62). At the same time, women in some developing countries are prone to high BMI due to issues such as excessive nutrition during pregnancy and postpartum weight retention (63, 64), thereby increasing their risk of cancer. In contrast, the primary burden in men stems from colon and rectal cancer, liver cancer, and kidney cancer. The age distribution indicates that individuals aged 50–79 bear the major burden, which aligns with the age-related characteristics of cancer incidence (risk increases with age) (65, 66). However, the rising trend among younger populations (aged 20–24) is concerning, possibly reflecting the early impact of increasing obesity rates on cancer incidence in younger individuals. Therefore, cancer prevention and control measures should be tailored based on gender and age differences in cancer burden. First, establish gender-specific screening guidelines, such as breast cancer screening for women and liver cancer prevention for men. Second, implement a life-course weight management program, with targeted interventions during high-risk periods of weight fluctuation, such as pregnancy and menopause. Third, develop a youth obesity-cancer risk prediction model to initiate early interventions for individuals with persistent overweight.

Another innovative aspect of this study compared to previous research is the use of the ARIMA model to predict the burden of various cancers attributable to high BMI over the next decade. The burden of multiple cancers is projected to continue rising through 2031. This trend highlights the long-term carcinogenic effects of obesity as a chronic risk factor, while also being closely associated with factors such as population aging and increased consumption of processed foods (67, 68). Although the ASMR of colon and rectal cancer and uterine cancer show declining trends, their absolute burden may still increase given the persistent rise in obesity rates. These projections stress the persistent growth trend of high BMI-related cancer burden and highlight the urgency for future public health interventions. First, lifestyle interventions should be implemented for medium and low-risk populations, such as promoting low-sugar and low-fat diets to reduce calorie intake and minimize fat accumulation. Secondly, high-risk populations should undergo early cancer surveillance through standardized screening protocols (such as fecal occult blood tests, capsule endoscopy). Finally, for patients already diagnosed with cancer, a multidisciplinary team (MDT) approach should be adopted to enhance quality of life through integrated medical treatment, symptom management, psychological support, nutritional optimization, social care, and other supportive measures. However, implementing the above measures poses certain challenges. Potential challenges may include insufficient long-term public adherence, complexities in individualized implementation, and a lack of systemic and resource support. To address these challenges, we can implement tiered management strategies and promote the development of digital health tools (health monitoring apps and others).

This study also has several limitations. First, the GBD database relies on cancer registry data reported by individual countries, and statistical incompleteness due to various factors may lead to an underestimation of the disease burden. Second, during data extraction, colon cancer and rectal cancer were aggregated as “colon and rectal cancer,” while gallbladder cancer and extrahepatic bile duct cancer were combined under “gallbladder and biliary tract cancer.” These malignancies exhibit distinct variations in anatomical location, biological behavior, risk factors, and therapeutic approaches. Their aggregation may obscure significant differential trends among the constituent subtypes. This methodological limitation may compromise the accurate interpretation of disease burden dynamics and subsequently hinder the development of targeted prevention strategies. In addition, the ARIMA model does not account for unforeseen events or external shocks within its predictive framework. For instance, during pandemic lockdowns and social distancing policies, reduced outdoor activity and prolonged sedentary behavior, coupled with shifts in dietary patterns, may have exacerbated obesity rates, thereby elevating the associated disease burden (69).

This study systematically evaluates the causal relationship between BMI and 12 types of cancer, along with their global burden distribution patterns, through an integrated approach combining Mendelian randomization analysis and the GBD database. This study reveals that the global burden of cancers attributable to high BMI continues to exhibit a constant growth trend. The distribution of cancer burden and the proportional contribution of specific cancers are intricately modulated by factors such as economic regions, geographic areas, sex, and age. In the future, only through multidisciplinary collaboration, precision prevention, and global governance can we effectively curb the spread of this “silent pandemic” and alleviate its long-term threats to human health.