Introduction

High body mass index (BMI), a widely used indicator for measuring overweight and obesity, has become a global public health concern1. The prevalence of high BMI is rising globally, particularly among adolescents, posing a serious threat to their current and future health and well-being2. Worldwide adult obesity has more than doubled since 1990–2022, and adolescent obesity has quadrupled. 37 million children under the age of 5 were overweight, and over 390 million children and adolescents aged 5–19 years were overweight in 2022, including 160 million who were living with obesity3. Adolescence is a critical period of physical, psychological, and social development, and high BMI during this stage can have long-term detrimental consequences. Overweight and obese adolescents are at increased risk of developing a range of chronic diseases, including type 2 diabetes, hypertension, certain types of cancer, and musculoskeletal disorders4. These conditions not only impact their physical health but also impair their quality of life, leading to social stigma, reduced self-esteem, and limitations in daily activities5.

The economic burden associated with high BMI is substantial. The direct healthcare costs associated with treating obesity-related diseases, such as diabetes and cardiovascular disease, place a strain on healthcare systems globally6. Epidemiological trends indicate a concerning rise in high BMI prevalence among adolescents globally, with variations observed across different regions, sexes, and socioeconomic backgrounds7. While some high-income countries have experienced a plateau or slight decline in high BMI prevalence among adolescents in recent years, low- and middle-income countries, particularly those undergoing rapid urbanization and socioeconomic transitions, are witnessing a rapid increase in prevalence8. This highlights the need for tailored interventions that address the unique challenges faced by different populations.

While previous studies have documented the rising prevalence of adolescent obesity, a critical examination of existing global policies aimed at mitigating this trend is warranted. International organizations and national governments have implemented various strategies, ranging from educational campaigns promoting healthy lifestyles to regulatory measures targeting the food industry9. However, the effectiveness of these policies varies considerably across contexts. For instance, Swinburn et al. (2019) highlight the potential of tax and regulatory policies on ultra-processed foods to reduce childhood obesity, particularly in settings where these foods are widely accessible and affordable10. Furthermore, addressing the social determinants of health, such as the built environment and access to healthy, affordable food options, is crucial11,12. Urban environments characterized by a high density of fast-food outlets and limited opportunities for physical activity can exacerbate the risk of obesity, particularly among socioeconomically disadvantaged populations13. A comprehensive approach that integrates policy interventions with efforts to address the broader social and environmental factors influencing adolescent health is essential.

Previous studies have explored the burden of high BMI among adolescents who often relied on data from specific countries or regions, limiting the generalizability of findings to the global context. Additionally, some studies focused on specific age groups within the adolescent range, hindering a comprehensive understanding of trends across the entire 10–19 year age group14. Moreover, there is a need for more recent data to reflect the evolving landscape of high BMI globally. This study aims to provide a comprehensive analysis of the global epidemiology of high BMI among adolescents aged 10–19 years, utilizing data from the Global Burden of Disease (GBD) Study. We aim to: (1) analyze the global, regional, and national prevalence of high BMI and its trends from 1990 to 2021; (2) investigate disparities in high BMI prevalence across sex, nation, and socio-demographic index (SDI). By elucidating trends, geographical variations, and demographic disparities, this research seeks to inform the development of targeted interventions and public health strategies to mitigate the growing burden of high BMI among adolescents.

Our findings reveal that global prevalence of high BMI among adolescents aged 10–19 years rises from 8.36% in 1990 to 17.64% in 2021, with females showing slightly higher rates than males. High SDI countries exhibit the highest prevalence, while middle SDI countries experience the fastest increases. Significant geographic variations persist, with rapid rises in regions like the Pacific Island nations and slower growth or declines in parts of East Asia. The Bayesian age-period-cohort model predicts a continued increase in global prevalence to 20.88% by 2030, with notable differences across countries.

Methods

Data sources

This study utilized publicly available data from the GBD Study 2021, obtained via the Global Health Data Exchange (GHDx) (https://ghdx.healthdata.org/gbd-results-tool). The GBD Study aggregates and anonymizes data from various sources, such as surveys, censuses, and administrative records, to generate comprehensive estimates of health loss worldwide15. All data used in this analysis were fully anonymized and aggregated at the population level, making individual identification impossible. Therefore, informed consent from individual participants was not required for this secondary analysis. The GBD Study itself adheres to strict ethical protocols. The collection and processing of data within the GBD framework were approved by the Institutional Review Board (IRB) of the University of Washington, which waived the need for individual informed consent for the aggregated, de-identified data used in GBD analyses. Further details regarding GBD’s ethical standards and data governance can be found on the official GBD website. Consequently, no additional specific ethical approval or individual informed consent was required for the present study, which solely relied on these publicly accessible, de-identified GBD estimates.

The 2021 iteration included data for 204 countries and territories, categorized into 21 geographically defined regions and further aggregated into seven super-regions based on shared mortality patterns16. Analyses were stratified by sex (male, female, and both sexes combined) across 23 age groups, and by SDI, a composite measure of socioeconomic development that incorporates lag-distributed income per capita, average years of schooling, and fertility rates among women younger than 25 years17. The GBD Study 2021 includes countries and territories based on data availability and quality, assessed via criteria encompassing completeness, accuracy, and representativeness. Statistical adjustments are employed to enhance comparability across diverse data sources. While aiming for global coverage, data availability, and quality vary, with high-income countries generally having more robust data than low-income countries18. Detailed information on data sources, inclusion/exclusion criteria, and quality assessment procedures is publicly available through the GHDx website.

GBD Study data is publicly accessible and can be downloaded without the need for specific approval or application. We downloaded the relevant data from the GBD Results Tool available on the GHDx website (https://ghdx.healthdata.org/gbd-results-tool). This tool allows users to select specific diseases or risk factors (high BMI in this study), age groups (10–19 years), sexes (male, female, and both combined), and time ranges (1990–2021). When downloading the data, we selected the ‘Location’ option, which included all countries and territories; the ‘Measure’ option, selecting the ‘Prevalence’ indicator; the ‘Cause’ option, selecting “High Body Mass Index”; and the ‘Age Group’ option, selecting 10–19 years. We downloaded the datasets containing prevalence estimates and their corresponding 95% uncertainty intervals (UIs). These data were downloaded in CSV format and used for subsequent statistical analyses. The downloaded datasets included 204 countries and territories, 23 age groups, and data stratified by sex and SDI.

High BMI in the GBD 2021 study, specifically for children and adolescents aged 2–19 years, is defined using the International Obesity Task Force (IOTF) standards for overweight and obesity14. The IOTF BMI cut-offs, derived from a six-country reference population, provide age- and sex-specific standards for ages 2–18 years, linked to adult BMI values at key thresholds (16, 17, 18.5, 25, and 30 kg/m²)19. These age-corresponding cut-offs are detailed in the Supplementary data 1.

Descriptive analysis

To characterize the global burden of high BMI among adolescents, we first examined the prevalence of high BMI in 1990 and 2021. Prevalence estimates were calculated for both sexes combined and stratified by sex (female and male). To assess disparities in the burden of high BMI, we further stratified prevalence estimates by SDI for five levels: high, high-middle, middle, low-middle, and low. We then explored geographical variations in high BMI prevalence by analyzing prevalence estimates in 1990 and 2021 across 7 super regions and 204 countries and territories. This analysis focused on identifying countries that exhibited particularly high or low prevalence rates in 2021, as well as those with the most rapid or substantial changes in prevalence over time.

Trend analysis

We employed estimated annual percentage change (EAPC) to quantify the temporal trends in high BMI prevalence among adolescents aged 10–19 years globally. The EAPC represents the average annual rate of change in prevalence over the study period20. Assuming a linear relationship between the natural logarithm of the prevalence rate (ln[prevalence]) and the calendar year, EAPC was calculated using linear regression models21. Specifically, the equation ln(prevalence) = α + β* year + σ was used, where β represents the regression coefficient and σ represents the error term. The EAPC was then derived using the formula: EAPC = 100*(exp(β)−1). A negative EAPC indicates a declining trend, while a positive EAPC indicates an increasing trend22. We calculated EAPCs to assess overall global trends, as well as trends stratified by sex (female and male), SDI, GBD region, and individual country.

To further investigate the temporal patterns of high BMI prevalence and identify statistical changes in trends, we employed joinpoint regression analysis23,24. This method models the trend as a series of joined linear segments, with each segment representing a period of constant percentage change. The points where the segments connect are referred to as “joinpoints” and indicate statistically significant shifts in the trend. For each segment, an annual percentage change (APC) was calculated using the formula: APC = [(yx+1−yx)/yx]*100 = (eβ1−1)*100, where β1 is the slope of the regression line obtained from the equation ln(y) = β0 + β1x, with y representing the prevalence and x representing the year. Joinpoint regression analysis was conducted using Joinpoint software (Version 4.9.1.0; Statistical Research and Applications Branch, National Cancer Institute, https://surveillance.cancer.gov/joinpoint/)25,26.

In the GBD Study, uncertainty is systematically quantified for all estimates to account for various sources of variation, including uncertainty in primary data sources, modeling approaches, data errors, and data manipulation27. This uncertainty is captured statistically in the 95% UIs reported for each location and estimate. UIs are calculated from 1000 draw-levels from the posterior distributions of the models, with the 2.5th and 97.5th percentiles defining the 95% UI range. Both prevalence and number estimates are presented with their corresponding 95% UIs.

Correlation analysis

To investigate the relationship between socioeconomic development and high BMI prevalence, we performed Spearman’s rank correlation analysis. This non-parametric method was used to assess the association between each country’s SDI in 2021 and both the high BMI prevalence in 2021 and EAPC of high BMI prevalence from 1990 to 2021. Scatter plots were generated to visualize these correlations, with Spearman’s rank correlation coefficients (ρ) and corresponding p-values included to quantify the strength and statistical significance of the relationships.

Predictive analysis

To project future trends in high BMI prevalence, we employed a Bayesian age-period-cohort analysis (BAPC) model28,29. This approach allows for the simultaneous estimation of the effects of age, period, and birth cohort on prevalence while accounting for the inherent uncertainty associated with each factor. We used the integrated nested Laplace approximation (INLA) method to fit the BAPC model, as it has been shown to provide more accurate and computationally efficient estimates compared to traditional Markov Chain Monte Carlo (MCMC) sampling techniques28. The INLA approach directly approximates the posterior distributions of the model parameters, reducing the risk of convergence issues and providing more reliable predictions. The BAPC model was used to predict global high BMI prevalence among adolescents up to the year 2030.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Results

Global trends in high BMI prevalence, 1990–2021

The global prevalence of high BMI among adolescents aged 10–19 years increased from 8.36% (95% UI, 8.02–8.69) in 1990 to 17.64% (95% UI, 16.91–18.36) in 2021, corresponding to an estimated 88,206,868 adolescents with high BMI in 1990 and 227,664,632 in 2021. This represents an EAPC of 2.4% (95% CI, 2.36–2.44) over the three decades. This upward trend was apparent in both sexes. Females had a slightly higher prevalence in both 1990 (8.63%, 95% UI, 8.18–9.11) and 2021 (18.35%, 95% UI, 17.33–19.42) compared to males (8.10%, 95% UI, 7.71–8.55 in 1990 and 16.97%, 95% UI, 16.11–17.91 in 2021). In 1990, an estimated 44,614,445 female and 43,589,084 male adolescents had high BMI globally. By 2021, these numbers had increased to 114,976,702 and 112,687,554, respectively. The EAPC for females was also slightly higher (2.44%, 95% CI 2.39–2.48) than for males (2.36%, 95% CI 2.32–2.41) (Fig. 1).

Fig. 1: Global prevalence and estimated annual percentage change (EAPC) of high BMI among adolescents (10–19 years), 1990–2021.
figure 1

a Prevalence of high BMI by sex in 1990 and 2021. b EAPC in the prevalence of high BMI by sex, 1990–2021. c Prevalence of high BMI by socio-demographic index (SDI) in 1990 and 2021. d EAPC in the prevalence of high BMI by SDI, 1990–2021.

Full size image

Socioeconomic disparities in high BMI prevalence

In 1990, the prevalence of high BMI among adolescents was 16.78% (95% UI, 15.85–17.78) in high SDI countries, 9.01% (95% UI, 8.49–9.48) in high-middle SDI countries, 7.76% (95% UI, 7.35–8.25) in middle SDI countries, 5.99% (95% UI, 5.51–6.51) in low-middle SDI countries, and 5.19% (95% UI, 4.83–5.56) in low SDI countries. These prevalence rates translated to an estimated 21,306,301, 16,776,917, 28,752,455, 15,420,243, and 5,866,725 adolescents with high BMI in each respective SDI group. By 2021, the prevalence in each SDI group had risen to 32.74% (95% UI, 31.06–34.22), 21.84% (95% UI, 20.73–22.85), 19.56% (95% UI, 18.50–20.59), 14.37% (95% UI, 13.30–15.46), and 10.35% (95% UI, 9.67–11.12), respectively, corresponding to 39,343,824, 32,958,212, 73,448,382, 54,310,163, and 27,441,648 adolescents with high BMI. The EAPC from 1990 to 2021 was 2.12% (95% CI, 1.97 to 2.27) for high SDI, 2.93% (95% CI, 2.88–2.98) for high-middle SDI, 3.02% (95% CI, 2.99–3.06) for middle SDI, 2.85% (95% CI, 2.83–2.88) for low-middle SDI, and 2.24% (95% CI, 2.18–2.30) for low SDI countries (Fig. 1).

Regional and national trends in high BMI prevalence, 1990–2021

Substantial geographical variations were observed in both the prevalence of high BMI among adolescents and its temporal trends, with some regions experiencing rapid increases while others showed slower growth or even declines (Figs. 2–4).

Fig. 2: Global distribution of high BMI prevalence (%) among adolescents (10–19 years) in 1990.
figure 2

This map illustrates the geographical distribution of high BMI prevalence among adolescents aged 10–19 years in 1990. Warmer colors (red and orange) indicate a higher prevalence of high BMI, while cooler colors (green and blue) indicate a lower prevalence of high BMI.

Full size image
Fig. 3: Global distribution of high BMI prevalence (%) among adolescents (10–19 years) in 2021.
figure 3

This map illustrates the geographical distribution of high BMI prevalence among adolescents aged 10–19 years in 2021. Warmer colors (red and orange) indicate a higher prevalence of high BMI, while cooler colors (green and blue) indicate a lower prevalence of high BMI.

Full size image
Fig. 4: Estimated annual percentage change (EAPC) in high BMI prevalence (%) among adolescents (10–19 years), 1990–2021.
figure 4

This map depicts the estimated annual percentage change (EAPC) in high BMI prevalence among adolescents aged 10–19 years globally between 1990 and 2021. Warmer colors (red and orange) indicate a higher annual increase in prevalence, while cooler colors (green and blue) indicate slower increases.

Full size image

High-income countries

Among high-income countries in 2021, the United States of America had the highest prevalence of high BMI among adolescents (42.32%, 95% UI, 38.86–45.83), followed by New Zealand (39.54%, 95% UI, 35.05–44.66), Greenland (36.92%, 95% UI, 30.62–42.83). The most rapid increases in prevalence from 1990 to 2021 were observed in Singapore (EAPC 3.83%, 95% CI 3.73–3.93), Brunei Darussalam (EAPC 3.47%, 95% CI 3.42–3.52), and the Republic of Korea (EAPC 2.71%, 95% CI 2.51–2.91). In contrast, the lowest prevalence rates were found in Japan (17.75%, 95% UI, 14.33–21.57), Republic of Korea (19.13%, 95% UI, 15–23.12), Switzerland (19.58%, 95% UI, 15.92–23.41), Andorra (24.99%, 95% UI, 20.2–29.66), and Sweden (24.33%, 95% UI, 20.23–28.94).

Central Europe, Eastern Europe, and Central Asia

In 2021, the highest prevalence of high BMI among adolescents in Central Europe, Eastern Europe, and Central Asia was observed in Montenegro (29.75%, 95% UI, 24.66–34.57), followed by Georgia (27.3%, 95% UI, 22.65–31.97), Albania (22.24%, 95% UI, 17.99–27.12), Azerbaijan (22.73%, 95% UI, 18.78–28.04), and Croatia (23.25%, 95% UI, 18.44–27.91). The most rapid increases in prevalence from 1990 were in Uzbekistan (EAPC 4.17%, 95% CI 4.12–4.23) and Serbia (EAPC 2.44%, 95% CI 2.39–2.49). Conversely, the lowest prevalence rates were found in Ukraine (15.50%, 95% UI, 12.71–18.72) and Slovakia (15.53%, 95% UI, 12.52–19.34).

Latin America and Caribbean

In 2021, the highest prevalence of high BMI among adolescents in Latin America and the Caribbean was observed in the United States Virgin Islands (45.44%, 95% UI, 39.64–51.24), followed by Puerto Rico (43.01%, 95% UI, 37.14–48.78), Mexico (42.23%, 95% UI, 37.43–47.32), Bermuda (40.19%, 95% UI, 34.83–46.35), and Bahamas (38.05%, 95% UI, 32.11–44.00). The fastest increases in prevalence were noted in Guatemala (EAPC 5.97%, 95% CI 5.89–6.04), Peru (EAPC 4.71%, 95% CI 4.62–4.80). The lowest prevalence rates were found in Haiti (12.27%, 95% UI, 9.62–15.55) and Honduras (18.26%, 95% UI, 14.51–22.30).

North Africa and Middle East

In 2021, the highest prevalence of high BMI among adolescents in North Africa and the Middle East was found in the United Arab Emirates (59.21%, 95% UI, 53.19–64.73), followed by Kuwait (57.03%, 95% UI, 51.23–62.58), Saudi Arabia (52.60%, 95% UI, 47.00–58.60), Bahrain (52.81%, 95% UI, 46.03–58.94), and Qatar (53.97%, 95% UI, 47.64–59.96). Oman had the most rapid increase in prevalence, with an EAPC of 3.71% (95% CI 3.67–3.76), followed by the Syrian Arab Republic (EAPC 3.72%, 95% CI 3.68–3.76), Iran (Islamic Republic of) (EAPC 3.84%, 95% CI 3.81–3.86), Tunisia (EAPC 3.62%, 95% CI 3.61–3.63), and Algeria (EAPC 3.28%, 95% CI 3.20–3.36). The lowest prevalence rates were observed in Yemen (18.68%, 95% UI, 14.90–22.79), Afghanistan (16.07%, 95% UI, 12.89–19.75), Sudan (25.93%, 95% UI, 21.30– 30.68), Morocco (24.65%, 95% UI, 19.92–29.61), and Lebanon (45.02%, 95% UI, 38.66–51.47). The most substantial declines in prevalence were noted in Qatar (EAPC −1.53%, 95% CI −1.61 to −1.44), followed by the United Arab Emirates (EAPC −1.22%, 95% CI −1.31 to −1.13), and Saudi Arabia (EAPC −1.20%, 95% CI −1.3 to −1.1).

South Asia

In 2021, the highest prevalence of high BMI among adolescents in South Asia was observed in Bhutan (14.36%, 95% UI, 11.34–17.95), followed by Pakistan (10.94%, 95% UI, 8.46–13.70), Nepal (10.13%, 95% UI, 8.09–12.90), and India (9.92%, 95% UI, 7.64–12.08). The fastest increase in prevalence from 1990 was in India (EAPC 2.98%, 95% CI 2.86–3.11), followed by Nepal (EAPC 2.78%, 95% CI 2.72–2.83). The lowest prevalence rate was found in Bangladesh (7.22%, 95% UI, 5.42–9.25).

Southeast Asia, East Asia, and Oceania

In 2021, the highest prevalence of high BMI among adolescents in Southeast Asia, East Asia, and Oceania was observed in the Cook Islands (74.02%, 95% UI, 68.53–78.85), followed by Nauru (72.03%, 95% UI, 66.13–77.02) and Tonga (71.14%, 95% UI, 66.33–75.89). The most rapid increases in prevalence from 1990 were seen in China (EAPC 3.71%, 95% CI 3.67–3.75), Maldives (EAPC 3.59%, 95% CI 3.56–3.63), Timor-Leste (EAPC 3.42%, 95% CI 3.36–3.48). Conversely, the lowest prevalence rates were reported in Cambodia (5.20%, 95% UI, 4.05–6.67), and Viet Nam (6.20%, 95% UI, 4.76– 7.79).

Sub-Saharan Africa

In 2021, the highest prevalence of high BMI among adolescents in Sub-Saharan Africa was observed in Equatorial Guinea (41.87%, 95% UI, 36.34–48.87), followed by Comoros (30.70%, 95% UI, 25.41 to 36.94), Gabon (26.23%, 95% UI, 21.58– 31.19), South Africa (25.94%, 95% UI, 22.37–30.58). The most substantial increases in prevalence from 1990 were observed in Ethiopia (EAPC 5.45%, 95% CI 5.33–5.56), Cameroon (EAPC 4.85%, 95% CI 4.76–4.93), Zambia (EAPC 4.48%, 95% CI 4.38–4.59). The lowest prevalence rates were seen in Niger (6.14%, 95% UI, 4.61–7.89), Eritrea (7.16%, 95% UI, 5.57–9.14), Burundi (7.66%, 95% UI, 5.76–9.80).

Tropical Latin America

In 2021, the highest prevalence of high BMI among adolescents in Tropical Latin America was seen in Puerto Rico (43.01%, 95% UI, 37.14–48.78), followed by Mexico (42.23%, 95% UI, 37.43–47.32), Bermuda (40.19%, 95% UI, 34.83–46.35), Belize (35.88%, 95% UI, 30.18–41.91). The most rapid increases in prevalence from 1990 were observed in Guatemala (EAPC 5.97%, 95% CI 5.89–6.04), Peru (EAPC 4.71%, 95% CI 4.62–4.80), Ecuador (EAPC 3.48%, 95% CI 3.41–3.56). Conversely, the lowest prevalence rates were found in Haiti (12.27%, 95% UI, 9.62–15.55) and Suriname (17.36%, 95% CI 13.71–21.26).

Global trends in high BMI prevalence: joinpoint regression analysis by sex and SDI level

Joinpoint regression analysis revealed fluctuations in high BMI prevalence trends among adolescents globally between 1990 and 2021 (Fig. 5). Global prevalence increased from 1990 to 1994 (APC 2.9%), 1994–1999 (APC 2.7%), 1999–2004 (APC 2.5%), 2004–2009 (APC 2.4%), 2009–2016 (APC 2.1%), and 2016–2021 (APC 2.3%). In females, the prevalence increased from 1990 to 1994 (APC 2.9%), 1994–1999 (APC 2.7%), 1999–2005 (APC 2.6%), 2005–2009 (APC 2.5%), 2009–2017 (APC 2.1%), and 2017–2021 (APC 2.3%). Among males, the prevalence increased from 1990 to 1994 (APC 2.8%), 1994–1999 (APC 2.7%), 1999–2003 (APC 2.4%), 2003–2008 (APC 2.3%), 2008–2016 (APC 2.1%), and 2016–2021 (APC 2.4%).

Fig. 5
figure 5

Trends in the prevalence of high BMI among adolescents (10–19 years) globally, 1990–2021, by sex (joinpoint regression analysis).

Full size image

In high SDI countries, prevalence increased from 1990 to 1995 (APC 3.4%), 1995–1999 (APC 3.3%), 1999–2002 (APC 2.6%), 2002–2005 (APC 2.2%), 2005–2009 (APC 1.7%), and 2009–2021 (APC 1.4%). In high-middle SDI countries, prevalence increased from 1990 to 1994 (APC 3.1%), 1994–2004 (APC 2.5%), 2004–2007 (APC 3.1%), 2007–2014 (APC 3.4%), 2014–2017 (APC 3.0%), and 2017–2021 (APC 2.5%). In low SDI countries, prevalence increased from 1990 to 1996 (APC 1.7%), 1996–2000 (APC 2.0%), 2000–2011 (APC 2.2%), 2011–2015 (APC 2.5%), 2015–2019 (APC 2.8%), and 2019–2021 (APC 3.0%). In low-middle SDI countries, prevalence increased from 1990 to 1997 (APC 2.8%), 1997–2004 (APC 3.1%), 2004–2008 (APC 2.8%), 2008–2014 (APC 2.6%), 2014–2019 (APC 2.8%), and 2019–2021 (APC 3.3%). In middle SDI countries, prevalence increased from 1990 to 1994 (APC 3.6%), 1994–1998 (APC 3.1%), 1998–2004 (APC 2.9%), 2004–2010 (APC 3.3%), 2010–2013 (APC 3.0%), and 2013–2021 (APC 2.6%) (Fig. 6).

Fig. 6
figure 6

Trends in the prevalence of high BMI among adolescents (10–19 years) globally, 1990–2021, by SDI (joinpoint regression analysis).

Full size image

Correlation between socioeconomic development and high BMI prevalence among adolescents

Spearman’s rank correlation analysis revealed a positive correlation between a country’s SDI in 2021 and the prevalence of high BMI among adolescents in the same year (R = 0.54, p < 0.001) (Fig. 7A). This suggests that higher levels of socioeconomic development are associated with a higher prevalence of adolescent high BMI. In contrast, a weak negative correlation was observed between SDI and the EAPC of high BMI prevalence from 1990 to 2021 (R = −0.23, p = 0.00091) (Fig. 7B). This indicates that countries with higher SDI tended to have a slower average annual increase in high BMI prevalence over the study period, although the relationship was not as strong as that observed for prevalence in 2021.

Fig. 7: Correlation between socio-demographic Index (SDI) and high BMI prevalence among adolescents.
figure 7

a Correlation between SDI in 2021 and high BMI prevalence in 2021. b Correlation between SDI in 2021 and the estimated annual percentage change (EAPC) of high BMI prevalence from 1990 to 2021.

Full size image

Predictive analysis of high BMI prevalence to 2030

To project future trends, a BAPC model was utilized, predicting global high BMI prevalence among adolescents (10–19 years) up to 2030 (Fig. 8). The model suggests a continued upward trajectory in prevalence. Globally, the prevalence is projected to rise from 17.23% in 2021 to 20.88% in 2030. In females, prevalence is predicted to increase from 17.80% in 2021 to 21.21% in 2030, while in males, it is projected to rise from 16.70% to 20.50% over the same period.

Fig. 8: Projected global prevalence of high BMI among adolescents (10–19 years), 1990–2030, by sex (Bayesian age-period-cohort analysis).
figure 8

a Overall prevalence. b Prevalence by sex. Prevalence from 1990 to 2021 is based on observed data. Projections from 2022 to 2030 are generated using a Bayesian age-period-cohort analysis model. Shaded areas represent 95% uncertainty intervals.

Full size image

BAPC analysis was also used to project high BMI prevalence in 2030 for each individual country (Fig. 9). The projections reveal substantial variation in the anticipated burden of high BMI among adolescents across the globe. Nauru is projected to have the highest prevalence of high BMI in 2030 (82.57%), followed by the United Arab Emirates (77.40%) and the Cook Islands (75.35%). Several other Pacific Island nations, including Tonga, Kiribati, and Samoa, are also projected to have prevalence rates exceeding 60%. In contrast, the lowest projected prevalence rates are found in Georgia (2.01%), Pakistan (8.31%), and Vietnam (7.95%).

Fig. 9: Projected prevalence of high BMI (%) among adolescents (10–19 years) in 2030.
figure 9

This map illustrates the projected prevalence of high BMI among adolescents aged 10–19 years in 2030, based on a Bayesian age-period-cohort analysis model. Darker shades of blue indicate a higher projected prevalence, highlighting anticipated regional variations in the burden of high BMI.

Full size image

Discussion

This study provides a comprehensive analysis of the global, regional, and national prevalence of high BMI among adolescents aged 10–19 years from 1990 to 2021, as well as projections for 2030. Our findings reveal an increase in high BMI prevalence globally over the past three decades, with a consistent upward trend. Females exhibit a slightly higher prevalence compared to males, although both sexes show similar increases over time. Notable disparities exist across SDI levels. While high SDI countries have the highest prevalence in 2021, middle SDI countries experienced the most rapid annual increase from 1990 to 2021. Substantial geographical variations are evident in both prevalence and temporal trends, with some regions, such as the Pacific Island nations, experiencing particularly rapid increases, while others, like parts of East Asia, show slower growth or even declines. Looking ahead, the BAPC model projects a continued rise in global high BMI prevalence among adolescents up to 2030, with considerable variation anticipated across individual countries.

The alarming rise in high BMI prevalence among adolescents globally is a major public health concern, reflecting a complex interplay of dietary, behavioral, socioeconomic, and cultural factors. Increased consumption of ultra-processed foods, sugary drinks, and unhealthy fats, often readily available and heavily marketed, has been strongly linked to weight gain and obesity in adolescents30,31. Simultaneously, sedentary lifestyles, characterized by increased screen time, reduced physical activity in schools, and limited opportunities for active leisure, are contributing to an energy imbalance that favors weight gain32. Socioeconomic factors play a crucial role, as poverty, food insecurity, and limited access to healthy food options create environments where unhealthy dietary choices become the default33,34. Urban environments, with their abundance of fast-food outlets and limited access to safe spaces for physical activity, further exacerbate these challenges35. Cultural factors, including changing dietary patterns and norms, as well as the pervasive influence of marketing and advertising that promotes unhealthy foods and beverages, contribute to the normalization of obesogenic behaviors36.

As illustrated by our analysis, the prevalence and trends of high BMI among adolescents vary considerably across different regions, underscoring the complex interplay of region-specific factors. These variations emphasize the need for tailored interventions and public health strategies that consider the unique contexts of each region. Cultural norms around food and body size can play a role in shaping dietary habits and perceptions of healthy weight37. However, with the globalization of food systems and the influx of processed foods, these traditional views may now be contributing to a higher acceptance of overweight and obesity, coupled with less emphasis on physical activity38. In contrast, many East Asian cultures place a greater emphasis on healthy eating and physical activity, which may contribute to the relatively lower prevalence of high BMI observed in these regions39. Economic development is another key factor influencing dietary patterns and activity levels. As countries undergo rapid economic growth, they often experience nutrition transitions characterized by increased consumption of energy-dense, processed foods, and decreased physical activity40. This pattern is evident in many middle-SDI countries, where we observe the fastest increases in high BMI prevalence despite lower baseline levels in 1990. The rapid urbanization accompanying economic development often leads to changes in built environments, with less access to safe spaces for physical activity and a greater density of fast-food outlets13.

The particularly rapid rise in high BMI prevalence among adolescents in middle-SDI countries is of particular concern. These countries are undergoing a complex transition, marked by rapid economic growth, urbanization, and lifestyle changes8. This transition often leads to a shift towards more sedentary lifestyles, increased consumption of processed foods, and a decline in traditional dietary patterns38. Moreover, there may be a lag between economic development and the adoption of public health policies and programs that promote healthy lifestyles. As a result, these countries face a dual burden of both undernutrition and overnutrition, with high BMI prevalence increasing rapidly alongside persistent challenges of undernutrition and micronutrient deficiencies41. Early intervention is crucial in middle-SDI countries to prevent a future epidemic of obesity-related diseases. Comprehensive strategies are needed, addressing both individual behaviors and the broader environmental and social determinants of health. These strategies should include promoting healthy diets, increasing physical activity levels, regulating food environments, and strengthening healthcare systems to provide early identification and management of overweight and obesity42.

Examining successful public health interventions in different countries provides valuable insights for addressing the global rise in adolescent obesity. Fiscal policies, such as taxes on sugar-sweetened beverages (SSBs), have shown promise in reducing consumption and, consequently, mitigating weight gain. Grummon and Hall (2020) highlight the potential benefits of SSB taxes, while also acknowledging the challenges and recommending strategies for effective implementation43,44. Countries like Mexico and Chile, which have implemented SSB taxes, have reported reductions in SSB purchases and consumption45,46. Beyond taxation, other effective policies include comprehensive school-based programs that combine nutrition education, physical activity promotion, and healthy food provision, as well as regulations restricting the marketing of unhealthy foods and beverages to children9. These examples demonstrate that multi-pronged, evidence-based policy interventions are crucial for creating environments that support healthy choices.

The COVID-19 pandemic has further exacerbated the global challenge of adolescent obesity, with widespread lockdowns and school closures leading to disruptions in daily routines and lifestyle behaviors. Increased sedentary behavior due to reduced opportunities for physical activity and increased screen time, coupled with changes in eating patterns, including increased consumption of processed foods and snacks, have contributed to weight gain in many adolescents47. Rundle documented the negative impact of social isolation during the pandemic on physical activity levels among adolescents, highlighting the importance of social interaction and structured activities for maintaining healthy habits48. The pandemic’s long-term consequences on adolescent obesity are still unfolding, but it is clear that the pandemic has underscored the need for resilient public health systems and strategies that can effectively address the complex interplay of factors contributing to obesity, even during times of crisis.

Beyond socioeconomic disparities, cultural norms and political decisions shape the obesogenic environment. Cultural preferences for certain foods, traditional dietary practices, and societal views on body image can influence eating behaviors and physical activity levels. Moreover, local policies, particularly those regulating the marketing of unhealthy foods and beverages to children, play a crucial role. As Boyland et al. demonstrated, exposure to advertising for unhealthy food is strongly associated with increased intake, highlighting the impact of the commercial food environment49. Furthermore, specific vulnerable populations, such as ethnic minorities and adolescents in rural areas, often face compounded challenges, including limited access to affordable, healthy food options and fewer opportunities for safe physical activity13. Addressing these multifaceted influences requires a comprehensive approach that considers both individual behaviors and the broader socio-cultural and political context.

Furthermore, emerging evidence suggests that biological mechanisms may contribute to the increased susceptibility to obesity observed in socioeconomically disadvantaged populations. One such mechanism is the “Thrifty Phenotype” hypothesis, which posits that individuals exposed to nutritional deprivation in utero or during early infancy may develop metabolic adaptations that favor energy storage50,51. While these adaptations may have been advantageous in environments with limited food availability, they can become maladaptive in settings with abundant access to energy-dense foods, leading to an increased risk of obesity and related metabolic disorders. It is possible that current adolescents from lower socioeconomic backgrounds, even if not directly exposed to undernutrition, may inherit this programming from previous generations. Another potential biological determinant is the dysregulation of the Hypothalamus–Pituitary–Adrenal (HPA) axis. Chronic exposure to psychosocial stress, which is often more prevalent in disadvantaged communities, can lead to sustained elevations in cortisol levels (hypercortisolism)52,53. Hypercortisolism, in turn, has been linked to increased appetite, preferential deposition of visceral fat, and a heightened risk of developing addictive behaviors, including those related to food consumption54,3. These biological pathways provide a plausible explanation for the observed disparities in obesity prevalence, highlighting the complex interplay between environmental stressors, physiological responses, and individual behaviors.

Our analysis consistently shows a slightly higher prevalence of high BMI among female adolescents compared to males, although both sexes experience a similar upward trend over time. This finding aligns with previous research on adolescent obesity, which has documented sex differences in BMI trajectories55. However, understanding whether these differences are primarily biologically driven or shaped by social factors requires careful consideration. Biological factors, such as hormonal differences, body composition, and potential genetic predispositions, may contribute to the observed sex disparities. Estrogen, the primary female sex hormone, plays a role in fat deposition and distribution, potentially leading to a higher body fat percentage in females compared to males56. Genetic factors may also influence susceptibility to obesity, and some studies have suggested that certain genes associated with obesity risk may be expressed differently in males and females57,58. Social factors, however, are likely to exert a greater influence on the observed sex differences. Gender norms around food and body image can shape dietary behaviors and attitudes towards weight. In many societies, girls and women face greater pressure to conform to thin ideals, leading to unhealthy dieting behaviors and body dissatisfaction59. Moreover, gender roles and expectations often limit girls’ opportunities for physical activity compared to boys. Girls may be discouraged from participating in certain sports or have less access to safe spaces for physical activity60,61. This disparity in physical activity can contribute to an energy imbalance, favoring weight gain in girls. Understanding the complex interplay of biological and social factors that contribute to sex differences in high BMI is crucial for developing effective and tailored interventions. Public health programs should address both individual behaviors and the broader social and cultural norms that influence these behaviors. Promoting healthy body image, encouraging physical activity for all genders, and creating supportive environments that foster healthy lifestyle choices are essential for reducing obesity among both boys and girls.

The BAPC model projects a continued rise in high BMI prevalence among adolescents globally, with variation across countries, carrying alarming implications for the future health and well-being of young people and highlighting the urgent need for effective interventions and policy action. If these trends continue unabated, we can anticipate a surge in the burden of chronic diseases associated with obesity, including type 2 diabetes, cardiovascular disease, certain types of cancer, and musculoskeletal disorders2. These conditions not only lead to reduced quality of life and increased disability but also contribute to premature mortality62. The projected increases in high BMI will inevitably strain healthcare systems, particularly in low- and middle-income countries already facing resource constraints, as the economic burden of treating obesity-related diseases is substantial and will likely escalate if these trends persist6. These projections should serve as a call to action for governments, public health agencies, and healthcare providers to prioritize adolescent obesity prevention and management. Comprehensive and evidence-based interventions are crucial, addressing both individual behaviors through education and support programs promoting healthy diets and physical activity, and the broader environments that shape those behaviors9. Policy action is essential to create supportive environments, including regulating food environments by implementing policies to limit the marketing of unhealthy foods to children, improving food labeling, and making healthier options more accessible; promoting physical activity by investing in safe and accessible spaces, encouraging active transportation, and incorporating physical activity into school curricula; and addressing socioeconomic inequalities by tackling poverty and food insecurity, ensuring equitable access to healthcare and education, and creating opportunities for economic advancement. A multi-sectoral approach involving collaboration between government, industry, communities, and individuals is essential to effectively address the complex challenges posed by adolescent obesity. By acting now, we can strive to reverse these concerning trends and ensure a healthier future for young people around the world.

While our BAPC model provides valuable insights into potential future trends in adolescent obesity, it is crucial to acknowledge the inherent uncertainties associated with such projections. As Foreman et al. emphasize, forecasting future health outcomes based on past trends is subject to various limitations63. The model’s projections rely on the assumption that past trends will continue in a similar manner, which may not hold true due to unforeseen changes in social, economic, environmental, or political factors. Furthermore, the model’s accuracy is dependent on the quality and completeness of the available data, which, as previously discussed, can vary across countries and regions. The 95% uncertainty intervals presented alongside our projections provide a measure of statistical uncertainty, but they do not capture all sources of potential error. Therefore, these projections should be interpreted as likely scenarios under the assumption of continued trends, rather than definitive predictions of future prevalence. It is also important to consider alternative scenarios that deviate from the assumption of continued trends. The implementation of more aggressive and effective public health policies, such as widespread adoption of taxes on sugar-sweetened beverages, stricter regulations on food marketing to children, and substantial investments in promoting physical activity, could alter the projected trajectory of adolescent obesity9. Conversely, unforeseen events, such as economic downturns, major global conflicts, or future pandemics, could also have potentially negative impacts on adolescent health and well-being. While our model does not explicitly quantify these alternative scenarios, it is essential to recognize that the future prevalence of adolescent obesity is not predetermined and will be shaped by the actions and policies implemented in the coming years.

This study benefits from the use of data from the GBD Study 2021, a comprehensive and rigorously conducted global effort to quantify health loss, applying a standardized methodology and drawing upon a wide range of data sources to ensure consistency and comparability across countries and time periods. The GBD Study’s large sample size enhances the generalizability of our findings, and our analysis covers a long time period (1990–2030), allowing for a thorough examination of temporal trends and the projection of future trajectories. Furthermore, our application of advanced statistical techniques, including BAPC analysis and the INLA method, enhances the accuracy and robustness of our predictions while accounting for uncertainty. However, limitations include our reliance on secondary data, which constrains us from the available variables and potential data quality variations between countries. Additionally, we were unable to explore all potential risk factors for high BMI due to data availability. Future research should prioritize improving data collection and quality globally, expanding the range of risk factors studied, and conducting longitudinal studies to gain a more nuanced understanding of the factors driving adolescent obesity.

Despite the strengths of our study, including the use of comprehensive data from the GBD Study 2021 and advanced statistical methods, several limitations warrant consideration. As highlighted by the NCD Risk Factor Collaboration (NCD-RisC, 2020), standardizing anthropometric data collection across diverse populations presents challenges64. While the GBD Study employs rigorous methods to harmonize data from various sources, including adjusting for differences in measurement techniques and data quality18, residual biases may still exist. Variations in BMI measurement methods, the availability and quality of health records, and cultural differences in reporting health information can introduce inconsistencies across countries. Furthermore, while we provide a general description of data sources in the GBD study in the methods section, it should be noted that detailed explanations of the criteria for including and excluding countries and populations analyzed are publicly available on the GBD official website. Although the GBD study has been updated, the criteria used in selecting countries and populations for analysis were based on data availability and quality. The GBD collaborators make concerted efforts to include as many countries and territories as possible, data from some regions may be more limited or less reliable than others, potentially impacting the generalizability of our findings to all populations. Future research should prioritize improving data collection and standardization efforts, particularly in low- and middle-income countries, to enhance the accuracy and comparability of global health estimates.

Conclusion

The global surge in adolescent high BMI prevalence, projected to continue rising, demands immediate, concerted action. Disparities across regions, socioeconomic levels, and sexes necessitate tailored interventions, but concrete policy changes are paramount. Governments must prioritize comprehensive strategies, including fiscal policies like taxes on sugar-sweetened beverages and subsidies for healthy foods; strict regulations on marketing unhealthy foods to children; comprehensive school-based programs integrating nutrition and physical activity; urban planning that promotes active living; and clear front-of-package food labeling. Simultaneously, international organizations like the WHO should set global standards, provide technical assistance, monitor trends, and mobilize resources. Addressing socioeconomic inequalities, promoting public awareness, and improving access to healthy options are crucial. Only through such a multi-faceted, collaborative approach, encompassing governments, international bodies, and communities, can we create environments that empower adolescents to make healthy choices and reverse this alarming epidemic, securing a healthier future.