Epidemiological trends of burden of brain and CNS cancer at global, regional, and national level: a trend analysis study from 1990 to 2021

Abstract

Background

Brain and central nervous system (CNS) cancers, though relatively rare, contribute substantially to neurological disability and cancer-related mortality worldwide. This study systematically assessed their global, regional, and national trends from 1990 to 2021 and projected burden through 2035 using Global Burden of Disease Study (GBD) 2021 data.

Methods

Using data from the Global Burden of Disease (GBD) 2021 study, we estimated age-standardised incidence rate (ASIR), prevalence rate (ASPR), death rate (ASDR), and DALY rate across 204 countries, 21 regions, and five socio-demographic index (SDI) groups. Trends were analyzed using estimated annual percentage change (EAPC), and future projections were generated using a Bayesian age-period-cohort (BAPC) model.

Results

From 1990 to 2021, the global ASIR of brain and CNS cancers increased from 3.75 (95% UI 3.21–4.21) to 4.28 (95% UI 3.71–4.88) per 100,000, with an EAPC of 0.45 (95% CI 0.40–0.49). ASPR rose from 8.66 (95% UI 7.55–9.53) to 12.01 (95% UI 10.54–13.52), while ASDR remained stable, and DALY rates declined from 119.88 (95% UI 99.23–137.57) to 107.91 (95% UI 91.74–125.59). High-income Asia Pacific showed the largest ASIR increase (EAPC 2.02), while Central Asia experienced the highest rise in DALYs (EAPC 1.20). Age and sex disparities were notable, with older adults (≥ 70 years) and males exhibiting higher rates. Projections indicate continued ASIR and ASPR growth, while ASDR and DALY rates decline, and–critically–an increase in the absolute numbers of cases, deaths, and DALYs through 2035 driven by population growth and ageing.

Interpretation

Brain and CNS cancers show increasing incidence and prevalence globally, with stable mortality and declining DALYs. These trends highlight the need for enhanced early detection, equitable access to care, and targeted interventions across regions. Notwithstanding broadly stable or declining age-standardised rates, absolute counts of cases and deaths are projected to rise to 2035, underscoring the urgency of scaling diagnostic and treatment capacity—particularly in lower-SDI settings.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40001-025-03511-5.

Introduction

Brain and central nervous system (CNS) cancers, although relatively uncommon, pose a substantial public-health challenge owing to their aggressive behaviour, complex histopathology, and often poor prognosis [1, 2]. Most arise in the brain–over 90% of cases—but tumours also occur in the spinal cord, meninges, and cranial nerves [1, 3]; within the WHO framework, approximately 30.2% are malignant and 69.8% benign [4]. Beyond high morbidity and neurological sequelae, limited therapeutic options translate into considerable disability and health-system burden [5, 6]. In 2020, an estimated 308,102 incident cases and 251,329 deaths were attributable to brain and CNS cancers worldwide [7]. Despite their lower frequency relative to many other cancers, these tumours contribute disproportionately to cancer-related disability and mortality [8]. Marked geographic heterogeneity persists, with incidence, mortality, and disability adjusted life years (DALYs) varying widely across regions and countries [9]. Over recent decades, age-standardised incidence and prevalence have increased, whereas age-standardised mortality and DALY rates have remained broadly stable [1, 10], underscoring the need for reliable, comparable, and longitudinal data to inform policy and planning.

Prior analyses have often been restricted to single countries, select years, or specific populations, limiting their relevance to global strategies [8, 11]. Leveraging the Global Burden of Disease (GBD) 2021 study, we estimate the burden of brain and CNS cancers across 204 countries and territories from 1990 to 2021, disaggregated by age, sex, and socio-demographic index (SDI). We further project trends to 2035 using a Bayesian age–period–cohort model (BAPC).

Methods

Data sources and case definition

This study obtained estimates from the GBD 2021 study via the Global Health Data Exchange (GHDx; Institute for Health Metrics and Evaluation), a comprehensive, systematic analysis of health metrics covering 204 countries and territories across 21 GBD regions [12, 13]. Data were accessed on 8 May 2025. We analysed the GBD cause brain and CNS cancer (B.1.24; cause-id 477), comprising malignant primary CNS tumours of the brain, spinal cord, meninges, and cranial nerves, mapped primarily to ICD-10 C70–C72 and ICD-9 191–192. Non-malignant and behaviour-unspecified CNS tumours are not included, and recurrent or secondary (metastatic) tumours are excluded, so only primary malignant cases are analysed. In GBD, case ascertainment draws on population-based cancer registries, vital registration, hospital records, mortality surveillance, and the published literature; diagnoses are identified from histopathology, imaging (eg, MRI or CT), or clinical records where histology is unavailable. GBD compiles and standardises these inputs using validated methods, and age-standardised incidence (ASIR), prevalence (ASPR), death (ASDR), and DALY rates are reported using the GBD standard population, enabling comparability across locations and time.

Socio-demographic index

The SDI is an aggregate measure employed within the GBD framework to reflect regional disparities in development and health outcomes [14]. It is derived as the geometric mean of three core indicators: per capita income, average educational attainment among individuals aged 15 years and above, and the total fertility rate among women younger than 25 years. Based on their SDI values, countries and territories were classified into five categories: low, low-middle, middle, high-middle, and high SDI regions. This classification facilitated a stratified analysis of the burden and trends of brain and CNS cancers by SDI, allowing for a more nuanced understanding of health disparities across different levels of socio-economic development.

Bayesian age-period-cohort model

The BAPC model was used to assess temporal trends and project future disease burden for brain and CNS cancers. The BAPC model is a sophisticated statistical framework that decomposes the disease burden into three fundamental components [15]:

Age Effect ($\alpha$): Reflects age-specific differences in disease risk, capturing the impact of biological and demographic factors.

Period Effect (β): Represents temporal changes that affect all age groups simultaneously, accounting for shifts in healthcare, public awareness, and diagnostic practices.

Cohort Effect (γ): Captures generational differences, indicating how exposure to risk factors varies among birth cohorts.

The BAPC model is mathematically expressed as:

$$Y_{\mathit{ijk}}=\mu +{\alpha }_i+{\beta }_j+{\gamma }_k+{ϵ}_{\mathit{ijk}}$$

where $Y_{\mathit{ijk}}$ represents the observed disease burden for age group $i$, period $j$, and cohort $k$; is the overall intercept; and ${ϵ}_{\mathit{ijk}}$ is the random error term.

Parameter estimation was performed using Markov Chain Monte Carlo (MCMC) methods, ensuring robust distributional estimates for age, period, and cohort effects. The convergence of MCMC chains was evaluated using the Gelman-Rubin diagnostic, enhancing model reliability.

Statistical analysis

Age-standardised rate (ASR) per 100,000 population were calculated using a weighted average approach, where each age group’s rate was adjusted based on the corresponding standard population proportion [12]. Specifically, the ASR was determined using the formula:

$$\text{ASR}=\frac{{\sum }_{i=1}^N{a}_i{w}_i}{{\sum }_{i=1}^N{w}_i}$$

where $a_i$ is the age-specific rate in the i-th age group, and $w_i$ represents the number of people in the same age group among the GBD standard population, and N is the number of age groups.

Temporal trends in ASR were evaluated using the estimated annual percentage change (EAPC). This was achieved by fitting a linear regression model to the natural logarithm of ASR:

$$ln\text{(ASR)}=\alpha +\beta \times \text{year}+\in$$

where α and β are the intercept and slope, respectively, and ϵ is the error term.

The EAPC was then calculated as:

$$\text{EAPC}=100\times (e^{\beta }-1)$$

The 95% confidence intervals (CIs) for EAPC were derived from the regression model, and trends were considered statistically significant if the CIs did not include zero.

To explore the relationship between ASR and the SDI, smoothing spline models were employed, providing a flexible approach to capture non-linear associations. Additionally, Spearman correlation analysis was used to assess the strength and direction of the association between ASR and SDI, with a significance threshold of P < 0.05.

All statistical analyses and visualizations were performed using R software (version 4.1.0), with a two-tailed -value of less than 0.05 indicating statistical significance.

Results

Global burden of brain and CNS cancers, 1990–2021

From 1990 to 2021, the global ASIR increased by about 14% (from 3.75 to 4.28 per 100,000) (Table 1). This rise was observed in both sexes, with a slightly faster increase in females than in males (EAPC 0.47 vs 0.42) (Table 1). Over the same period, the ASPR increased by about 39% (from 8.66 to 12.01 per 100,000), again with a modestly greater gain in females than in males (EAPC 1.28 vs 1.09) (Supplementary Table S1). In contrast, the ASDR remained essentially stable (from 3.04 to 3.06 per 100,000), reflecting a slight increase in males and near-stability in females (Supplementary Table S2). The age-standardised DALY rate declined by about 10% (from 119.88 to 107.91 per 100,000), with broadly similar patterns by sex (Supplementary Table S3).

Table 1.

The number of incidence cases and the ASIR of brain and CNS cancers in 1990 and 2021, and its temporal trends from 1990 to 2021

	1990		2021		1990–2021
	Number of cases (95%UI)	ASIR/100000(95% UI)	Number of cases (95%UI)	ASIR/100000(95% UI)	EAPC (95%CI)
Global (both)	173,086 (147,452, 194,951)	3.75 (3.21, 4.21)	357,482 (310,457, 407,433)	4.28 (3.71, 4.88)	0.44 (0.40, 0.49)
Global (male)	93,549 (73,313, 114,995)	4.18 (3.32, 5.05)	190,130 (148,118, 232,469)	4.72 (3.68, 5.76)	0.42 (0.38, 0.47)
Global (female)	79,537 (67,859, 92,584)	3.36 (2.91, 3.85)	167,353 (147,253, 187,695)	3.88 (3.43, 4.35)	0.47 (0.43, 0.51)
High SDI	55,694 (54,173, 56,847)	5.66 (5.52, 5.78)	100,994 (94,964, 105,629)	6.38 (6.08, 6.64)	0.43 (0.36, 0.50)
High-middle SDI	49,615 (42,242, 55,311)	4.81 (4.10, 5.35)	97,818 (83,423, 112,505)	5.90 (5.05, 6.77)	0.68 (0.63, 0.74)
Middle SDI	47,058 (35,595, 56,517)	3.35 (2.57, 4.07)	106,918 (86,956, 128,833)	4.11 (3.34, 4.94)	0.67 (0.65, 0.69)
Low-middle SDI	15,847 (11,648, 21,806)	1.73 (1.31, 2.32)	39,604 (31,858, 49,414)	2.34 (1.89, 2.93)	1.04 (0.99, 1.09)
Low SDI	4669 (2949, 7298)	1.23 (0.77, 1.75)	11,810 (8190, 15,144)	1.43 (1.00, 1.82)	0.46 (0.37, 0.55)
Andean Latin America	838 (665, 1161)	2.76 (2.20, 3.73)	2794 (2222, 3479)	4.43 (3.52, 5.50)	1.70 (1.40, 2.01)
Australasia	1393 (1343, 1448)	6.32 (6.10, 6.56)	2536 (2342, 2730)	5.95 (5.54, 6.38)	-0.14 (-0.22, -0.06)
Caribbean	897 (824, 1131)	2.93 (2.72, 3.61)	1971 (1725, 2270)	3.85 (3.37, 4.46)	1.25 (1.13, 1.37)
Central Asia	1966 (1640, 2244)	3.24 (2.68, 3.69)	4636 (4013, 5300)	4.94 (4.29, 5.62)	1.54 (1.41, 1.68)
Central Europe	8003 (7657, 8532)	5.76 (5.50, 6.15)	11,721 (10,689, 12,787)	6.64 (6.07, 7.25)	0.55 (0.31, 0.79)
Central Latin America	2825 (2740, 2916)	2.23 (2.17, 2.29)	7586 (6775, 8533)	2.98 (2.66, 3.36)	0.71 (0.45, 0.98)
Central Sub-Saharan Africa	295 (213, 419)	0.83 (0.60, 1.07)	876 (586, 1182)	1.03 (0.67, 1.39)	0.80 (0.68, 0.93)
East Asia	48,578 (35,372, 60,360)	4.63 (3.39, 5.77)	107,614 (83,225, 135,768)	6.02 (4.69, 7.53)	0.82 (0.78, 0.86)
Eastern Europe	9101 (8400, 9678)	3.63 (3.37, 3.85)	14,293 (13,134, 15,494)	4.98 (4.60, 5.38)	0.95 (0.86, 1.04)
Eastern Sub-Saharan Africa	1848 (1234, 2725)	1.25 (0.81, 1.67)	4821 (3457, 6278)	1.53 (1.05, 1.92)	0.70 (0.64, 0.75)
High-income Asia Pacific	5649 (5037, 5978)	3.11 (2.76, 3.29)	16,368 (14,108, 18,001)	5.44 (4.76, 5.92)	2.02 (1.73, 2.32)
High-income North America	22,170 (21,558, 22,637)	7.11 (6.93, 7.24)	36,462 (34,376, 37,686)	7.08 (6.75, 7.31)	0.02 (-0.05, 0.09)
North Africa and Middle East	10,151 (7656, 14,315)	4.09 (3.16, 5.72)	26,566 (19,734, 32,479)	4.98 (3.70, 6.10)	0.85 (0.75, 0.94)
Oceania	32 (16, 43)	0.69 (0.35, 0.94)	82 (43, 110)	0.74 (0.38, 0.99)	0.26 (0.19, 0.32)
South Asia	13,985 (9078, 18,699)	1.61 (1.05, 2.12)	31,817 (26,009, 43,157)	1.87 (1.53, 2.55)	0.34 (0.24, 0.43)
Southeast Asia	7051 (5029, 8699)	1.97 (1.43, 2.45)	16,691 (12,170, 19,950)	2.40 (1.76, 2.87)	0.67 (0.57, 0.77)
Southern Latin America	1485 (1350, 1622)	3.12 (2.84, 3.41)	2981 (2776, 3173)	3.81 (3.55, 4.06)	1.13 (0.84, 1.42)
Southern Sub-Saharan Africa	569 (441, 729)	1.57 (1.20, 1.98)	1390 (1024, 1668)	2.05 (1.50, 2.43)	0.91 (0.82, 1.00)
Tropical Latin America	4967 (4716, 5251)	4.14 (3.93, 4.40)	14,101 (13,434, 14,673)	5.65 (5.37, 5.88)	1.09 (0.86, 1.32)
Western Europe	30,521 (29,740, 31,118)	6.55 (6.40, 6.66)	49,619 (46,606, 51,779)	7.44 (7.13, 7.71)	0.48 (0.37, 0.60)
Western Sub-Saharan Africa	762 (515, 997)	0.43 (0.30, 0.53)	2558 (1315, 3276)	0.63 (0.34, 0.79)	1.47 (1.38, 1.56)

CNS central nervous system, ASIR age-standardised incidence rate, EAPC estimated annual percentage change

Regional and national burden of brain and CNS cancers, 1990–2021

From 1990 to 2021, global ASIR and ASPR of brain and CNS cancers increased across all SDI levels, while ASDR and DALYs showed a declining trend (Fig. 1, Supplementary Figure S1-S3). High-income Asia Pacific exhibited the highest increase in ASIR (EAPC 2.02, 95% UI 1.71–2.33), while a slight decline was recorded in Australasia (EAPC –0.14, 95% UI –0.23 to –0.06) (Table 1). The most pronounced rise in ASPR was observed in high-middle SDI regions (EAPC 1.91, 95% UI 1.82–1.99), whereas the lowest increase occurred in low SDI regions (EAPC 0.72, 95% UI 0.60–0.84) (Supplementary Table. S1). East Asia demonstrated the largest rise in ASPR, more than doubling from 9.63 (95% UI 6.94–11.67) per 100,000 in 1990 to 20.83 (95% UI 16.70–26.00) in 2021 (EAPC 2.62, 95% UI 2.55–2.69), while Oceania maintained the lowest ASPR throughout the period. For ASDR, the greatest increase was reported in low-middle SDI regions (EAPC 0.87, 95% UI 0.80–0.94), while high SDI regions experienced a modest decline (EAPC –0.15, 95% UI –0.21 to –0.09) (Supplementary Table. S2). In terms of DALYs, Central Asia had the highest increase (EAPC 1.20, 95% UI 1.05–1.35), whereas East Asia recorded the most significant decline (EAPC –1.07, 95% UI –1.17 to –0.96) (Supplementary Table. S3).

Fig. 1.

Trends in age-standardised incidence rate of brain and CNS cancers by sex and SDI levels, 1990–2021. CNS, central nervous system, SDI socio-demographic index

In 2021, substantial cross-country variation in CNS cancer burden was observed. For ASIR, Norway recorded the highest rate (14.96 per 100,000, 95% UI 13.93–16.07) with an EAPC of 0.68% (95% CI 0.29–1.07), while the lowest rates were observed in the Gambia and Mali (Fig. 1A, Supplementary Figure S4A). Regarding ASPR, Norway again ranked highest (98.34 per 100,000, 95% UI 91.47–106.14) with an EAPC of 1.25% (95% CI 0.73–1.76), followed by Finland (Fig. 1B, Supplementary Figure S4B). Conversely, the Gambia had the lowest ASPR (0.30 per 100,000, 95% UI 0.18–0.41, EAPC 0.68%, 95% CI 0.41–0.95). Montenegro exhibited the highest ASDR (7.77 per 100,000, 95% UI 6.11–10.25), while Mali and Niger recorded the lowest rates. Notably, Niger experienced a significant increase in ASDR (EAPC 2.49%, 95% CI 2.32–2.66) (Supplementary Figure S5-S6). For DALYs, Montenegro reported the highest rate (276.39 per 100,000, 95% UI 211.29–363.85), despite a declining trend (EAPC –0.37%, 95% CI –0.48 to –0.26), while the Gambia had the lowest burden (6.03 per 100,000) (Supplementary Figure S5-S6).

Age and gender disparities in the global burden of brain and CNS cancers

In 2021, the global burden of brain and CNS cancers displayed clear age-specific patterns across prevalence, incidence, DALYs, and mortality (Fig. 2). Rates were minimal in early life but began to rise sharply from age 35–39 years, reaching their highest levels in older age groups. Prevalence peaked among individuals aged 90–94 years, at 66.39 per 100,000 (95% UI 50.95–80.13) in males and 64.03 (95% UI 44.04–82.53) in females (Fig. 2B), with a slight decline thereafter. Incidence followed a similar pattern, peaking at 33.95 per 100,000 (95% UI 27.82–38.97) in males and 26.23 (95% UI 19.58–31.34) in females in the same age range (Fig. 2A). In contrast, age-standardised DALY rates peaked earlier, at age 50–54 years, reaching 208.95 per 100,000 (95% UI 158.93–266.26) in males and 144.86 (95% UI 127.70–168.16) in females (Fig. 2D). Mortality, measured by ASDR, was highest among individuals aged 70–74 years, with rates of 16.60 per 100,000 (95% UI 13.41–19.87) in males and 12.41 (95% UI 10.77–13.73) in females (Fig. 2C).

Fig. 2.

Trends in the number of incidence cases and ASIR (A), the number of prevalence cases and ASPR (B), the number of deaths and ASDR (C), and the number of DALYs and the age-standardised DALY rate (D) of brain and CNS cancers by age and gender globally. CNS central nervous system, ASIR age-standardised incidence rate, ASPR age-standardised prevalence rate, ASDR age-standardised death rate, DALYs disability adjusted life years

Sex differences were consistent across all metrics, with males experiencing higher rates than females, particularly in middle-aged and older populations. For instance, in the 85–89 age group, ASDR reached 23.07 per 100,000 (95% UI 18.46–26.43) in males compared to 16.42 (95% UI 13.05–18.82) in females (Fig. 2C). A similar male predominance was observed for DALYs from age 30–34 years onward, with 82.23 per 100,000 (95% UI 59.68–108.22) in males versus 57.06 (95% UI 48.71–64.99) in females (Fig. 2D).

Trends and correlations of brain and CNS cancers burden by socio-demographic index level

From 1990 to 2021, global ASIR, ASPR, ASDR, and age-standardised DALY rates for brain and CNS cancers increased across all SDI levels (Fig. 3, Supplementary Figure S7). Substantial geographical variation in the burden of brain and CNS cancers was observed, with disproportionately higher values in Central Europe and Tropical Latin America, where all metrics exceeded SDI-based expectations. In contrast, Southern and Western Sub-Saharan Africa consistently reported lower-than-expected rates. A positive correlation was noted between SDI and ASIR (r = 0.583, p < 0.001), ASPR (r = 0.821, p < 0.001), age-standardised DALY rate (r = 0.608, p < 0.001), and ASDR (r = 0.650, p < 0.001).

Fig. 3.

The ASIR (A) and ASPR (B) of brain and CNS cancers by 21 GBD regions and SDI, 1990–2021. CNS central nervous system, GBD Global Burden of Diseases, Injuries, and Risk Factors Study, SDI sociodemographic index, ASIR age-standardised incidence rate, ASPR age-standardised prevalence rate

In 2021, deviations from expected burden patterns were evident across 204 countries (Supplementary Figure S8-9). Norway, Denmark, and Monaco reported higher-than-expected ASIR and ASPR, while Montenegro, North Macedonia, and Bulgaria exhibited elevated DALYs and ASDR (Figs. 4 and 3). SDI was further confirmed to positively correlate with DALYs (r = 0.583, p < 0.001), ASDR (r = 0.625, p < 0.001), incidence (r = 0.742, p < 0.001), and prevalence (r = 0.803, p < 0.001), underscoring the significant influence of socio-economic factors on the burden of brain and CNS cancers.

Fig. 4.

The ASIR (A) and ASPR (B) per 100,000 population of brain and CNS cancers by country in 2021. CNS central nervous system, ASIR age-standardised incidence rate, ASPR age-standardised prevalence rate

Projected global burden of brain and CNS cancers by 2035

Projections from the global BAPC model indicate a modest increase in the ASIR and ASPR for brain and CNS cancers through 2035, while both ASDR and age-standardised DALY rates are expected to decline (Fig. 5). Specifically, the ASIR is projected to reach 5.99 per 100,000 (95% UI 5.68–6.30), and the ASPR is anticipated to rise to 16.19 per 100,000 (95% UI 16.19–17.54). In contrast, the age-standardised DALY rate is expected to decline to 127.85 per 100,000 (95% UI 120.31–135.39), and the ASDR is projected to decrease to 4.16 per 100,000 (95% UI 3.97–4.36).

Fig. 5.

The temporal trends of ASIR (A), ASPR (B), ASDR (C) and age-standardised DALY rate(D) of brain and CNS cancers from 1990 to 2021 and projections up to 2035. CNS central nervous system, ASIR age-standardised incidence rate, ASPR age-standardised prevalence rate, ASDR age-standardised death rate, DALYs disability adjusted life years

Despite divergent trends in age-standardised rates, the absolute burden of brain and CNS cancers is expected to rise substantially by 2035 (Supplementary Figure S10). The total number of incident cases is projected to reach 534,419 (95% UI 507,021–561,816), and the number of prevalent cases is expected to increase to 1,444,182 (95% UI 1,323,996–1,564,369). Similarly, total deaths from brain and CNS cancers are anticipated to rise to 371,167 (95% UI 353,748–388,587), and the total number of DALYs is projected to increase to 11,406,175 (95% UI 10,733,553–12,078,797) by 2035.

Discussion

This study provides a comprehensive analysis of the global, regional, and national burden of brain and CNS cancers from 1990 to 2021, revealing critical trends across age, sex, sociodemographic groups, and future projections. Our findings show increasing global ASIR and ASPR, alongside relatively stable ASDR and modest declines in age-standardised DALY rates, reflecting the complex interplay of demographic shifts, advances in diagnostic technology, and improvements in clinical management.

The increase in ASIR and ASPR is consistent with wider availability and uptake of MRI/CT, earlier detection, and stronger registry capture in high-income settings [16, 17]. In these high-resource contexts, systematic case ascertainment and public awareness have likely contributed to higher recorded prevalence [18].

Regional disparities vividly illustrate the profound influence of socio-economic factors on the burden of brain and CNS cancers. High-income regions, such as the Asia Pacific and high-middle SDI areas, have witnessed substantial increases in ASIR and ASPR. This upward trend is mainly attributed to their advanced diagnostic capabilities, well-established cancer registry systems, and longer life expectancies. These regions benefit from sophisticated healthcare infrastructure, including specialised neuro-oncology centers and interdisciplinary care teams, which enhance the detection and management of cancers[19, 20]. Interpretation in lower-SDI settings should account for data sparsity, diagnostic constraints, and incomplete vital statistics, which can attenuate observed rates and narrow apparent gradients [21, 22].

National-level analysis uncovers striking disparities in the burden of brain and CNS cancers. Norway and Finland consistently report among the highest ASIR and ASPR globally. Beyond the common factors shared by high-income regions, Norway has implemented a national cancer control plan that emphasizes early detection through regular health screenings and invests heavily in research and development for neuro-oncology [23, 24]. Finland, on the other hand, has a unique primary healthcare system that effectively coordinates referrals to specialized cancer care, ensuring timely diagnosis [25]. These tailored healthcare policies, combined with their robust infrastructure, enable comprehensive data capture in cancer registries, even for early and asymptomatic cases. In countries reporting low rates (e.g., the Gambia and Mali), limited access to neuroimaging, workforce shortages, and incomplete registry coverage likely contribute to under-ascertainment and may depress recorded incidence and prevalence [26–28].

Age and gender disparities in the burden of brain and CNS cancers are pronounced, with a sharp increase in incidence observed from ages 35–39 years, peaking in older populations. This age-dependent rise reflects the cumulative effects of genetic mutations, environmental exposures, and declining cellular repair mechanisms associated with ageing [29, 30]. The higher burden among older adults is consistent with the natural history of these cancers, which are more likely to develop with prolonged exposure to risk factors and age-related declines in immune surveillance [31]. The earlier peak in DALYs compared to mortality underscores the significant morbidity associated with brain and CNS cancers, which often lead to progressive neurological impairment, cognitive decline, and functional dependence [32]. This burden is particularly severe in regions with limited access to neuro-rehabilitation and palliative care [33].

Gender differences are also evident, with males consistently exhibiting higher rates of incidence, prevalence, DALYs, and mortality. Biologically, males may have a higher genetic susceptibility to CNS cancers, while hormonal differences, particularly the protective effects of estrogen in females, may reduce cancer risk in women [34, 35]. Behaviorally, men are more likely to engage in high-risk occupations involving exposure to carcinogens, such as industrial chemicals and ionising radiation [36, 37]. They are also more prone to lifestyle risk factors, including tobacco and alcohol use. These biological and behavioral factors directly contribute to the elevated cancer burden among males, highlighting the need for targeted prevention strategies.

Socio-demographic analysis confirms a positive correlation between SDI and ASIR, ASPR, and DALYs, reflecting the influence of socio-economic development on CNS cancer burden [9, 12]. High-SDI regions experience higher incidence and prevalence due to better diagnostic capacity and extended life expectancy, while low-SDI regions face a hidden burden, characterized by lower reported incidence but higher mortality and disability. These patterns should be interpreted in light of data-quality limitations–particularly in settings with sparse primary data–so that cross-regional contrasts are viewed as indicative rather than definitive. Gender intersects significantly with these socio-economic factors. In high-SDI areas with advanced healthcare infrastructure and greater health awareness, the gender gap in cancer detection and treatment outcomes may be reduced as both men and women have better access to services. However, in low-SDI regions, where healthcare resources are scarce, gender disparities are exacerbated. Limited access to care and cultural barriers may further impede women's ability to seek timely diagnosis and treatment, while men's higher exposure to occupational risks due to economic necessity remains a persistent factor contributing to their higher cancer burden. This complex interplay underscores the need for integrated approaches that consider both socio-economic development and gender-specific needs to improve cancer control globally [38–40].

Estimates for settings with sparse population-based cancer registries, limited pathology services, and incomplete vital statistics are necessarily more model-dependent; as a result, true burdens in low-SDI regions may be underestimated, and cross-regional contrasts may be biased toward well-resourced systems [21, 22]. In the GBD framework, uncertainty intervals reflect posterior variation in modelled draws, but cannot fully compensate for systematic gaps in case ascertainment or misclassification [21]. Accordingly, we interpret SDI gradients and between-country rankings with caution, and we prioritise direction and relative magnitude over small absolute differences when data density is low. Policy-wise, this argues for investment in registry coverage, diagnostic capacity (imaging, histopathology, and molecular typing), and mortality certification to improve ascertainment and comparability over time [22, 27].

Future projections using the BAPC model indicate a significant upward trend in the absolute number of brain and CNS cancer cases, deaths, and DALYs by 2035. Despite relatively stable or declining age-standardised rates, population growth and ageing are projected to expand the pool of individuals at risk substantially. Low- and middle-income regions are expected to bear the brunt of this increase, as their healthcare systems already face significant strain. We note that these projections are model-dependent, particularly in data-sparse settings; thus, the projected increases in absolute counts should be considered conservative estimates. The projected rise in cases and deaths will likely outpace the current capacity of these regions to provide adequate diagnosis, treatment, and palliative care [40, 41, 43].

To address these projected challenges, targeted public health interventions are urgently required. Rather than generic screening, we emphasise timely detection through strengthened referral and diagnostic pathways. Expansion of diagnostic facilities—especially in resource-limited areas‐is crucial. Additionally, training more healthcare professionals in neuro-oncology can enhance the quality of care. Prioritizing equitable access to advanced treatment modalities, including neurosurgery, radiotherapy, and chemotherapy, is essential [27, 44, 45]. International collaborations play a vital role in this regard, offering financial support, technology transfer, and capacity-building initiatives to strengthen healthcare systems in low- and middle-income regions. Through these concerted efforts, the escalating burden of brain and CNS cancers can be mitigated, reducing the impact on patients and healthcare systems globally [44, 46].

This study has several strengths. It provides a comprehensive analysis of brain and CNS cancer burden using data from the GBD database, which ensures high-quality, consistent data across countries and regions. The use of the BAPC model allows for robust trend analysis and future projections, enhancing the reliability of the findings. Moreover, the study's focus on age, gender, regional, and national disparities provides a nuanced understanding of the disease's global impact.

However, this study also has limitations. Despite extensive coverage, under-reporting remains a concern in low- and middle-income settings, and reliance on cancer registries may underestimate the true burden where surveillance is weak. Our analyses follow the GBD cause hierarchy for brain and CNS, which captures malignant primary CNS tumours; consistent disaggregation by histology or behaviour is not available across locations and years, so subtype-specific trends could not be assessed. In data-sparse settings, estimates are more model-dependent, which may bias regional comparisons downward and compress gradients; uncertainty intervals cannot fully correct for systematic gaps, so gradients and country rankings should be interpreted cautiously. Finally, we did not model temporal shifts in risk-factor exposure, which may influence future trends.

Conclusion

In conclusion, this study provides critical insights into the global burden of brain and CNS cancers, highlighting significant disparities across regions, age, gender, and socio-demographic groups. The findings underscore the urgent need for targeted interventions to improve early detection, enhance treatment access, and reduce mortality and disability. Strengthening healthcare systems, expanding diagnostic infrastructure, and promoting international collaboration will be essential to mitigating the growing burden of these malignancies.

Author contributions

Minfeng Tong and Lin Chen: designed and conceived the study and critical revision of the manuscript for important intellectual content. Fan Yang: data acquisition. Fan Yang, Jianming Zhang and BinBin Ren: analyzed the data. Fan Yang and Qi Tu drawn the picture. Jianming Zhang: study supervision. Lin Chen: finished the manuscript. All authors reviewed and validated the manuscript.

Funding

This study was supported by Key Project of Jinhua City 2022–3-102, FJY2024-1-01，JY2023-2-03).

Data availability

All data are available within the Article and its Supplementary Materials.

Declarations

Ethical approval and consent to participate

Informed consent was waived for publicly available databases extracted from the Global Health Data Exchange (GHDx).

Competing interest

The authors declare no competing interests.