Autism spectrum disorder (ASD) encompasses a broad range of genetic neurobehavioral conditions, primarily characterized by developmental challenges in social communication and stereotypical, rigid, or repetitive behaviors
[1]. Since its identification in 1943, ASD has garnered increasing attention
[2]. It is a complex condition with an unclear etiology and no current cure, typically manifesting in infancy and often becoming evident by the age of three
[3]. While individuals with ASD generally have a normal life expectancy, they often require varying levels of lifelong care from their parents or caregivers
[4]. Beyond individual challenges, the impact of ASD extends to families, healthcare systems, communities, and society as a whole.
Between 1990 and 2019, the global prevalence of ASD increased from 20.3 million to 28.3 million cases, representing a 39.3% rise and corresponding to 4.3 million disability-adjusted life years (DALYs), a measure of overall disease burden
[5-6]. A key feature of this increase is the well-documented sex difference in prevalence, with males diagnosed approximately three to four times more frequently than females
[7]. Hypotheses for this disparity point to both biological and social factors. Biologically, sex-specific genetic and hormonal influences are believed to play a significant role in determining susceptibility to ASD
[8]. Socially, gender norms may shape how ASD is recognized and treated, often leading to underdiagnosis and insufficient support for females with ASD
[9]. Together, these biological and social influences contribute to the overall burden of ASD.
Understanding sex-based differences in ASD burden across diverse socioeconomic contexts and geographic regions is essential for designing effective and equitable interventions. These disparities are shaped by a complex interplay of biological and social determinants of health, both of which evolve with age
[10]. As global populations age, it becomes increasingly important to examine how sex differences in ASD manifest across different life stages, particularly in older age groups where research remains limited
[11-12]. Moreover, variations in cultural and societal factors across regions and socioeconomic strata may influence patterns of sex difference in ASD
[13-14], further complicating the global picture. Addressing these variations through targeted, evidence-based approaches is crucial for reducing inequities and promoting health equity in the management and prevention of ASD.
Several studies have examined the global epidemiology of ASD using data from Global Burden of Disease (GBD) study
[6, 15-17]. Yang, et al
[6] analyzed GBD 2019 data and found no significant changes in the age-standardized rate over time, with sociodemographic index (SDI)-related disparities persisting from 1990 to 2019. They concluded that the global burden of ASD has remained predominantly in males and disproportionately affects higher-SDI populations. In contrast, Solmi, et al
[17] emphasized the need for greater focus on female patients, noting that this group has historically been underrepresented in ASD epidemiological and clinical research. Li, et al
[16] conducted a bibliometric analysis to explore potential ASD risk factors in addition to GBD data. Moreover, another study
[15] highlighted the burden of ASD among children under 5 years old, underscoring the importance of addressing early-life disparities using GBD 2019 data.
A recent study
[18] has presented updated global estimates of ASD prevalence and health burden based on GBD 2021, an in-depth understanding of sex differences in ASD across age groups, socioeconomic strata, and geographic regions over the past and next 30 years remains limited. Using data from the GBD 2021 and its forecasts, this study aims to evaluate the long-term trends and global distribution patterns of sex differences in the burden of ASD and to project future changes, thereby providing evidence for ASD prevention and management.
1 Materials and methods
1.1 Ethics statements
All data used in this study were sourced from publicly accessible databases and adhered to relevant ethical standards and data usage permissions. The study did not involve any personally identifiable information,and thus, no additional ethical approval was required.
1.2 Data sources
We used estimates from GBD 2021 and GBD Foresight studies to quantify sex differences in the global burden of ASD across the lifespan. The GBD 2021 Results Database, publicly accessible through the GBD Collaborative Network website, reports the estimates of disease burdens (e.g., mortality and morbidity) from 1990 to 2021 for 371 diseases and injuries across 204 countries and territories. For ASD, the epidemiological systematic literature review was initially conducted for GBD 2017 and updated in April 2021 for the GBD 2021 study. The GBD foresight study provides cause-specific forecasts of mortality, years of life lost, years lived with disability, and DALYs by age and sex through 2050 for 204 countries and territories under 5 scenarios: 1 reference scenario (the most likely future, based on historical trends) and 4 alternative scenarios in which specific risk factors are eliminated by 2050 (Safer environment scenario, improved childhood nutrition and vaccination scenario, improved behavioral and metabolic risks scenario, and a combination of these 3 scenarios)
[19]. This study used the GBD Foresight reference scenario.
1.3 Methods
The GBD study provides a range of estimates for disease burdens by the Institute for Health Metrics and Evaluation (IHME, 2024). This study retrieved age-standardized, age-specific, and sex-specific GBD estimates to assess DALY rates (per 100 000 population) for males and females with ASD. DALY accounts for both the severity of a condition and its impact on quality of life, offering a comprehensive and comparable measure of disease burden across populations. One DALY represents the loss of one year of full health. We extracted age-standardized DALY rate (ASDR) and 95% uncertainty interval (
UI; representing the 2.5th and 97.5th percentiles) for ASD globally
[20], disaggregating them by sex, age group, region, and income group from 1990 to 2050.
To evaluate sex-based disparities, we calculated the absolute difference (AD) in DALYs by subtracting female DALYs from male DALYs and the relative difference (RD) by dividing the AD by the female DALY rate. Additionally, we assessed changes in DALY between 1990 and 2021 and between 2021 and 2050 by calculating relative change (RC) for each sex.
Subgroup analyses were conducted to explore specific variations. We categorized age groups based on the disease profile of ASD: <5 years (toddler/preschool children), 5-14 years (school-aged children), 15-49 years (young adults), 50-69 years (middle-aged adults), and ≥70 years (older adults)
[17, 21]. Furthermore, we analyzed data based on the World Bank income classifications (low, lower-middle, upper-middle, and high groups) and World Bank regions (Sub-Saharan Africa, East Asia, South Asia, North Africa & Middle East, Latin America & Caribbean, High-income North America, and Cetral Europe & Easter Europe & Central Asia).
Additionally, we extracted data for 5 SDI groups from 1990 and 2021. SDI, published by IHME, measures a country’s sociodemographic development based on income, education, and fertility, with 5 levels ranging from low to high (low, low-middle, middle, high-middle, and high). Furthermore, we explored the relationships between sex differences in the ASD burden and SDI, universal health coverage (UHC), and the gender inequality index (GII). UHC, provided by the World Health Organization (WHO), tracks progress in health service coverage using 23 indicators, with values ranging from 0 to 100
[22]. GII, published by the United Nations Development Programme, measures gender inequality in reproductive health, empowerment, and the labor market, with lower values indicating greater equality
[23].
1.4 Statistical analysis
All analyses were conducted using R software (version 4.3.1). We analyzed the ADs and RDs in DALY between females and males across age groups, socioeconomic strata, and geographic regions. We calculated the average annual percentage change (AAPC) and its 95% confidence interval (
CI) to describe temporal trends in sex differences in the ASD burden from 1990 to 2050 using the Joinpoint regression models. AAPC with a 95%
CI >0 indicates an increasing trend, an AAPC with a 95%
CI <0 indicates a decreasing trend, and an AAPC with a 95%
CI including 0 indicates a stable trend
[24].
We performed Pearson correlation tests to estimate the associations between ADs/RDs of ASD and GII, SDI, and UHC. The correlation coefficient (r2) and its 95% CI were reported. Polynomial curves were also modeled. A two-sided P value <0.05 was considered statistically significant.
2 Results
2.1 Sex differences in ASDR at global, regional, and national levels
Globally, the ASDR was 195.9 per 100 000 (95%
UI 133.7 to 274.9) for males and 92.4 (95%
UI 63.7 to 130.1) for females in 1990, yielding an AD of 103.5 per 100 000. By 2021 the ASDR increased to 199.8 per 100 000 (95%
UI 136.3 to 282.0) in males and 94.5 per 100 000 (95%
UI 64.5 to 133.0) in females (AD=105.3) (
Table 1). The ASDRs are projected to rise further by 2050, reaching 203.7 (95%
UI 138.7 to 286.4) in males and 99.7 (95%
UI 68.5 to 140.2) in females (AD=104.0). Despite these increases, the global RD remained between 1.0 and 1.1 over the 60 years.
Regionally, the largest sex differences were consistently observed in World Bank upper-middle and high-income countries, high-middle and high-SDI regions, and in high-income North America and East Asia across 1990, 2021, and 2050 (
Table 1). East Asia exhibited the highest RD, ranging from 1.8 to 2.0 across all 3 time points. National-level analyses yielded consistent findings were in Supplementary
Table 1 (
https://doi.org/10.57760/sciencedb.xbyxb.00099).
2.2 RCs in ASDR (1990—2021 and 2022—2050) by sex
From 1990 to 2050, the increase in ASD-related disease burden among males mainly occurred in 1990—2021, whereas for females, most of the rise is projected to occur between 2022 and 2050. Between 1990 and 2021, South Korea, Japan, Singapore, the United Kingdom, Equatorial Guinea, Spain, China, and the Maldives exhibited RCs in male ASDR>12 (Supplementary
Figure 1A, Supplementary Table 2;
https://doi.org/10.57760/sciencedb.xbyxb.00099). During the same period, female ASDR declined in over 80 countries (RC<0) (Supplementary
Figure 1C, Supplementary Table 2;
https://doi.org/10.57760/sciencedb.xbyxb.00099). From 2022 to 2050, female ASDR is projected to increase in nearly all countries except Venezuela (Supplementary
Figure 1D, Supplementary Table 2;
https://doi.org/10.57760/sciencedb.xbyxb.00099), with the largest increases for both sexes expected in African countries (Supplementary Figure
1B and
1D, Supplementary Table 2;
https://doi.org/10.57760/sciencedb.xbyxb.00099).
2.3 Sex disparities in age-specific DALY rates
In 2021, age-specific DALY rates were consistently higher in males than females. The highest rates were observed in the <5 years age group (males: 221.9 per 100 000; females: 112.8 per 100 000;
Figure 1). DALY rates decreased with age for both sexes, from 221.9 (<5 years) to 131.1 (≥70 years) in males and from 112.8 to 48.5 in females. Conversely, RDs increased steadily with age, reaching 1.7 in the ≥70 years group (Supplementary Table 3,
https://doi.org/10.57760/sciencedb.xbyxb.00099). Regionally, East Asia consistently exhibited the most pronounced disparities, with RDs ranging from 2.0 (<5 years) to 2.5 (≥70 years) (
Figure 1; Supplementary Table 3,
https://doi.org/10.57760/sciencedb. xbyxb.00099). Comparable patterns were seen in 1990 and are projected for 2050 (Supplementary Figure
2 and
3, Supplementary Table 3;
https://doi.org/10.57760/sciencedb. xbyxb.00099).
2.4 Trends in age-specific DALY rates disparities (1990—2050)
Globally, ADs in DALY rates decreased over time in the 15-49 years (AAPC=-0.057, 95%
CI -0.062 to -0.051,
P<0.001) and 50-69 years groups (AAPC=-0.052, 95%
CI -0.083 to -0.022,
P=0.001), but increased in the <5 years (AAPC=0.063, 95%
CI 0.057 to 0.070,
P<0.001), 5-14 years (AAPC=0.029, 95%
CI 0.022 to 0.035,
P<0.001) and ≥70 years (AAPC=0.114, 95%
CI 0.056 to 0.172,
P<0.001) from 1990 to 2050 (
Figure 2A; Supplementary Table 4-5,
https://doi.org/10. 57760/sciencedb.xbyxb.00099). Similarly, RDs declined in the <5 years (AAPC=-0.135,
P<0.001), 5-14 years (AAPC=-0.167,
P<0.001), and 15-49 years (AAPC=-0.185,
P<0.001), remained stable in 50-69 years (AAPC=-0.021,
P=0.250), and increased in the ≥70 years (AAPC=0.254,
P<0.001) (
Figure 2B; Supplementary Table 4-6,
https://doi.org/10.57760/sciencedb.xbyxb. 00099). Region- specific analyses showed similar patterns, with age groups declining in high SDI regions (Supplementary Figure 4; Supplementary Table 4-5;
https://doi.org/10.57760/sciencedb.xbyxb.00099) and RDs increasing in most regions except for middle and high SDI settings (Supplementary Figure 5, Supplementary Table 4-6;
https://doi.org/10.57760/sciencedb.xbyxb. 00099).
2.5 Correlations of sex differences with SDI, GII, and UHC
At the country level, both RD and AD were positively correlated with SDI (RD:
r=0.514,
P<0.001; AD: 0.508,
P<0.001; Figure
3A and
3B) and UHC (RD:
r=0.532,
P<0.001; AD:
r=0.517,
P<0.001; Figure
3C and
3D), but negatively correlated with GII (RD:
r=-0.574,
P<0.001; AD:
r=-0.634,
P<0.001; Figure
3E and
3F). Spain, Japan, Singapore, South Korea, and China displayed outlier levels of sex differences, deviating from the model-predicted trends (
Figure 3; Supplementary Table 7,
https://doi.org/10.57760/sciencedb. xbyxb.00099).
3 Discussion
To the best of our knowledge, this is the first study to use the most recent GBD data to analyze sex differences in DALY rates for ASD from 1990 to 2050. Consistent with earlier GBD estimates
[6, 16], our findings confirm that males consistently bear a higher ASD burden than females. This study extends previous findings by analyzing sex differences across different countries, regions, age groups, and trends over time, while also examining correlations with the SDI, UHC, and GII.
Our findings indicate that higher sex differences have consistently occurred in high-income and high-SDI regions from 1990 to 2050, with East Asia showing particularly pronounced disparities. In this region, the ASDR for ASD in males was more than twice that of females, with pronounced disparities across all age groups. These disparities may partly result from the exceptionally high prevalence of ASD in countries such as South Korea and Japan, where rates are particularly high in males
[25-26]. Moreover, deeply ingrained cultural norms, such as ‘son preference’
[11], may contribute to these disparities. This preference, rooted in profound social, economic, and cultural traditions in countries like China, Japan, and South Korea
[27], has led to imbalanced sex ratios
[28] and restricted access to rehabilitation and education for females with ASD
[29]. A study
[25] in South Korea further reinforces this, showing that in 2015, 87.5% of the total economic cost burden of ASD was associated with male patients, while only 12.5% was attributed to female patients
[25].
The analysis of sex-specific trends in ASD-related disease burden from 1990 to 2050 shows that while the increase in burden for males was concentrated between 1990 and 2021, females are projected to experience a more pronounced rise between 2021 and 2050. These rising trends are likely driven by factors such as improvements in awareness, detection and diagnosis of ASD, as well as greater financial investment in support services since ASD was first reported in 1943
[2]. Consequently, the number of individuals with ASD and the associated disease burden has surged
[30-31]. The difference in sex-specific trends may be due to improvements in the identification and support of females with ASD
[32-33], which could help reduce the sex-based disparity in growth trends in the future. Over the next 30 years, the growing ASD burden for both sexes is projected to be most pronounced in Africa. Despite progress in raising awareness and improving access to services, significant gaps remain in ensuring timely diagnosis, appropriate treatment, and adequate support for individuals with ASD in Africa
[34]. Addressing these challenges necessitates enhanced efforts in screening, diagnosis, and intervention to bridge these gaps and achieve better outcomes for individuals with ASD in the region.
Globally, the DALY rates for ASD were consistently higher in males than in females across all age groups in 2021, with the highest rates observed in children under 5 years for both sexes. This finding aligns with previous studies
[6, 21], reinforcing the importance of providing early support to children with ASD
[15, 17]. Interestingly, sex differences in DALY rates were most pronounced in individuals aged 70 and older, and this disparity is projected to continue growing from 1990 to 2050. The exact reasons for this pronounced disparity in the older age group are not fully understood, but several factors may contribute. First, the significant sex disparity among older individuals may stem from a large number of undiagnosed cases in females before the 1990s
[35]. Additionally, higher mortality rates among older females with ASD may contribute to the underrepresentation of their long-term disease burden
[36]. Second, aging-related disease burdens and health vulnerabilities, which are more pronounced in males
[8], could further exacerbate the disease burden for males. However, these results should be interpreted with caution, as they do not account for demographic shifts such as increasing life expectancy, declining fertility rates, and the longer lifespan of women
[37-38]. Given the expected growth of the elderly population and the widening sex differences, these findings underscore the need to consider both age- and sex-specific patterns in ASD care to ensure equitable support, particularly for the elderly.
Notably, sex differences in the ASD burden increased with higher SDI and UHC but decreased with greater GII. Higher SDI and UHC generally indicate greater awareness, improved understanding, and better access to healthcare services for ASD
[39-41]. However, this progress may not equally benefit both sexes. For instance, issues such as delayed and missed diagnoses, mismatched rehabilitation and educational support, and disproportionately economic investment continue to have larger influences on female than male ASD patients
[25, 42-43]. In contrast, the negative correlation between sex differences and GII may be misleading. This is because the GII primarily reflects gender-based disadvantage in areas such as reproductive health, empowerment, and the labor market
[44]. This may not fully capture the biological and social difference of ASD by sex. Furthermore, a lower GII signifies greater gender equality, which is typically associated with higher sociodemographic development, including improved access to healthcare and education
[45]. Therefore, the negative relationship between sex differences in the ASD burden and GII aligns with its positive association with SDI and UHC, reflecting broader trends in gender equality and health resource distribution. The exceptionally high sex differences in ASD burden observed in Spain, Japan, Singapore, South Korea, and China, which deviated from the model-predicted trends, may be attributed to factors such as low total fertility rates
[46], high ASD prevalence
[17], and cultural factors like son preference in these countries
[47-48].
Several limitations of this study warrant mention. First, the accuracy of GBD 2021 estimates depends on data availability and quality; data from low-income countries may be less reliable due to limited epidemiological studies and healthcare infrastructure, potentially leading to underestimation or misclassification of ASD cases. Second, variations in diagnostic methods across data sources and changes in diagnostic criteria over time (e.g., DSM-IV to DSM-5) may affect the consistency of burden estimates and observed trends. Third, this study did not analyze risk factors contributing to ASD burden, which warrants future investigation. Fourth, although this study identified significant sex differences in high-income/high-SDI regions, especially East Asia, attributing these solely to cultural factors like son preference is insufficient; further research is needed to understand these disparities.
In conclusion, sex differences in the burden of ASD persist globally and increase with age, highlighting the need for gender- and age-sensitive policies and interventions.