Mortality trends and disparities associated with aortic aneurysm and aortic dissection in the United States from 1999 to 2023: a CDC WONDER database analysis

Abstract

Background

Aortic aneurysm (AA) and aortic dissection (AD) are life-threatening cardiovascular diseases. AA includes thoracic (TAA) and abdominal (AAA) subtypes, while AD involves primary intimal tear. In the U.S., mortality trends, and demographic disparities remain unclear. This retrospective population-based study analyzed US AA/AD burden and disparities to guide targeted prevention and management.

Methods

We extracted AA/AD mortality data (1999–2023; urban-rural 1999–2020) for adults aged ≥ 25 in the continental U.S. from the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research, stratified by sex, age, race/ethnicity and region. We calculated crude and age-adjusted rates (CMR/AAMR), and estimated annual and average annual percentage change (APC/AAPC) via joinpoint regression. Two-tailed t-tests were used (p < 0.05).

Results

Between 1999 and 2023, there were 289,971 AA/AD-related deaths in the U.S., with AAMR declining (AAPC = -3.73%, 95% CI: -4.33 to -3.14; p < 0.001). Males and non-Hispanic Black adults had higher AAMR and slower declines. Crude mortality rose with age, while the 35–44 age group showed a small increase (AAPC = 1.04%, 95% CI: 0.63 to 1.46; p < 0.001). The Midwest exhibited a higher disease burden, and between 1999 and 2020, nonmetropolitan areas consistently showed greater disease burden compared with metropolitan regions.

Conclusions

From 1999 to 2023, U.S. mortality from aortic aneurysm and dissection declined overall, but burdens remained disproportionately high in non-Hispanic Black individuals, those aged 85+, and the Midwest. 1999–2020 data show greater burden in nonmetropolitan areas, highlighting persistent disparities in resources and risk management.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12872-026-05656-5.

Background

An aortic aneurysm (AA) is a localized aortic dilation caused by medial layer impairment from congenital/acquired factors and hemodynamic pressure [1]. It is classified into thoracic (TAA) and abdominal (AAA) subtypes by anatomical location [2]. Aortic dissection (AD) is characterized by a primary intimal tear that permits blood to enter the medial layer, creating a false lumen that may propagate in an anterograde or retrograde direction [3].The Stanford classification is widely used, with type A (corresponding to DeBakey I/II) involving the ascending aorta and carrying a poorer prognosis, and type B (DeBakey III/IIIb) affecting the descending aorta [4].

AA is a prevalent aortic disorder with substantial cardiovascular disease burden [5]. Global AAA incidence/prevalence has slightly declined, especially in high-income regions, linked to preventive measures like smoking cessation. TAA is mostly asymptomatic (incidence: 5–10/100,000 person-years) with an upward trend, possibly due to aging and improved imaging [6]. In the United States, the incidence and mortality associated with AAA have declined substantially and persistently; however, significant racial and geographic disparities exist, which are closely linked to economic development, healthcare access, and policy implementation [7]. However, over the past four decades, the incidence of TAA in the United States has increased more than threefold, partly attributed to population aging and wider adoption of chest imaging modalities. In contrast, mortality has declined gradually, with significant disparities observed across sex, ethnicity, and census regions [8].

Recent decades have witnessed advances in AA management, including conservative pharmacotherapy (risk factor control), open surgical repair, endovascular intervention (e.g., endovascular aneurysm repair, flow diversion devices), and emerging minimally invasive techniques [9–11]. Each strategy plays distinct roles: conservative treatment reduces rupture risk in low-risk unruptured cases (diameter <5 mm, asymptomatic) via standardized medication (e.g., β-blockers + statins) [12, 13]; Microsurgical clipping achieves high complete occlusion rates (≈ 94%) in wide-neck aneurysms and remains an effective option for complex aneurysms, particularly in selected younger patients where durable occlusion is desired [14].Endovascular approaches are favored for ruptured AA and elderly patients due to minimal invasiveness [15]; and novel flow diversion technologies improve outcomes of refractory cases [16].

AD is life-threatening, with TAA as a key underlying disease [4]. Its incidence is 2.6–3.5/100,000 person-years, driven by hypertension, aging, and atherosclerosis [6]. Meanwhile, AD exhibits marked geographical disparities, with patients in developing countries experiencing higher mortality rates due to diagnostic delays and inadequate treatment [17]. In the US, AD has a “dual trend”: reduced in-hospital mortality (especially type A) with advanced care [18], but stable incidence (3–5/100,000) linked to poor hypertension control, obesity, and substance abuse [6], plus, mortality disparities in elderly (> 85 years), males, and ethnic minorities [18].

Management of AD is guided by disease type. Acute type A AD requires urgent open surgical repair and remains the standard treatment, as early intervention markedly reduces short-term mortality [19, 20]. Uncomplicated type B AD is typically managed with optimal medical therapy, whereas thoracic endovascular aortic repair (TEVAR) or open surgery is indicated for complicated cases, including malperfusion or rapid aortic expansion [21, 22]. Recent registry data demonstrate an increasing utilization of TEVAR for type B AD; however, open surgical repair remains essential in selected complex cases, particularly those with extensive branch vessel involvement or unfavorable anatomy for endovascular intervention [21].

Despite advances in diagnosis and treatment, mortality trends and disparities in AA and AD remain incompletely defined. Analyses using the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research (CDC WONDER) database show that AA-related age-adjusted mortality in the U.S. declined from 1999 to 2020, yet significant disparities persist across sex, race/ethnicity, and region [8]. Similarly, AAA mortality in older adults varied by demographic groups [7]. Meanwhile, recent studies have shown that the overall mortality rate of AD in the US from 1999 to 2023 presented a decreasing trend, with significant geographical differences. In terms of age, the elderly population aged ≥ 85 years had the highest mortality rate; males consistently exhibited a higher mortality rate than females, and there were also differences in racial distribution [23]. Unlike prior single-disease studies, our dual-focus investigation of AA and AD uncovers their shared and distinct epidemiological features for more holistic risk stratification and targeted intervention.

This study analyzed the CDC WONDER data from 1999 to 2023, stratified by age, sex, race/ethnicity, region, and urbanization, to identify high-risk subgroups, temporal trends, and potential effects of management strategies. Metropolitan and nonmetropolitan data are available only through 2020, limiting assessment of urban–rural disparities across the full period. This research is clinically significant: it provides researchers with nationwide trend data, guides clinicians in tailoring management strategies over time, and informs patients about disease risk and prevention. We hypothesize that AA and AD mortality trends differ by sex, race/ethnicity, region, and urbanization status, potentially influenced by health policy, cultural factors, and socioeconomic conditions.

Methods

Data source

Data on AA/AD-related deaths among adults aged ≥ 25 years in the United States from 1999 to 2023 were retrieved from the CDC WONDER database (https://wonder.cdc.gov/mcd.html). Data stratified by urban–rural classification were available for the period 1999–2020. CDC WONDER is a comprehensive repository of death statistics covering all 50 U.S. states and the District of Columbia. This retrospective population-based study used the Underlying Cause of Death Public Use File to identify cases where AA or AD was documented as the underlying cause of death. Patients aged ≥ 25 years were identified using the International Classification of Diseases, 10th Revision (ICD-10) code I71 for aortic aneurysm and dissection. All subtypes of I71 were included in the study scope, Which classify AA and AD by anatomical location, rupture status, and dissection type: I71.0 (Dissection of aorta [any part]), I71.1 (Thoracic aortic aneurysm, ruptured), I71.2 (Thoracic aortic aneurysm, without mention of rupture), I71.3 (Abdominal aortic aneurysm, ruptured), I71.4 (Abdominal aortic aneurysm, without mention of rupture), I71.5 (Thoracoabdominal aortic aneurysm, ruptured), I71.6 (Thoracoabdominal aortic aneurysm, without mention of rupture), I71.8 (Aortic aneurysm of unspecified site, ruptured), and I71.9 (Aortic aneurysm of unspecified site, without mention of rupture). The study population consisted of individuals diagnosed with AA/AD and related diseases who died as a result of these conditions. This database focuses primarily on mortality data extracted from death certificates, providing key information regarding place of death and various patient-specific demographic characteristics, as well as cause of death and other relevant details. Additionally, this study included all deaths in the United States during a specific period in which AA/AD was listed as a potential or contributing cause of death. Exclusion criteria were applied to deaths with incomplete demographic data (i.e., missing values for any variable, particularly patient demographics) or unreliable reported data (i.e., inaccurate or untrustworthy data provided by patients/participants). This approach ensured that the analysis focused on reliable and relevant data, thereby enhancing the accuracy of the study findings. Since the CDC WONDER database contains de-identified and publicly available data, this study was exempt from institutional review board (IRB) approval.

Data extraction

We extracted data on AA- and AD-related mortality, corresponding population estimates, demographic variables (sex, age, race/ethnicity), and geographic information, including census region (1999–2023) and urban–rural classification (available for 1999–2020). Race and ethnicity were categorized into four groups: non-Hispanic (NH) White, NH Black, Hispanic, and NH Other. Age was stratified into the following groups: 25–34, 35–44, 45–54, 55–64, 65–74, 75–84, and 85 + years. Mortality trends were further evaluated across U.S. census regions (Northeast, Midwest, South, and West). Additionally, data were stratified by urban-rural classification in accordance with the 2013 National Center for Health Statistics (NCHS) Urban–Rural Classification Scheme, which comprises Nonmetropolitan (Micropolitan, NonCore) and Metropolitan (Large Central Metro, Large Fringe Metro, Medium Metro, Small Metro) categories [24]. A detailed CDC WONDER data extraction protocol is provided in the supplementary materials for reproducibility.

Statistical analysis

Calculation of crude and age-adjusted mortality rates

Crude Mortality Rates (CMR) and Age-Adjusted Mortality Rates (AAMR) per 100,000 population were calculated. CMR was derived by dividing the annual number of AA- and AD-related deaths by the corresponding annual U.S. population size. AAMR was computed by direct standardization to the 2000 U.S. standard population, which is defined by the CDC/NCHS as the age distribution of the U.S. population in the year 2000 across standard 5-year age groups (0–4, 5–9, …, ≥ 85 years) to allow comparison across time periods and populations [25].

Joinpoint settings

Temporal trends in AAMR and CMR were assessed using the Joinpoint Regression Program (version 4.9.0.0; National Cancer Institute, Bethesda, MD, USA). Joinpoint regression analyses were conducted using log-transformed mortality rates to estimate the annual percentage change (APC) and average annual percentage change (AAPC) and corresponding 95% confidence intervals. The maximum number of joinpoints was set to three, in accordance with the study period and to minimize overfitting. Model selection was performed using Monte Carlo permutation tests, with statistical significance determined at p < 0.05. Standard errors were calculated internally by the Joinpoint software assuming a Poisson distribution of deaths, and models were weighted by the inverse of the variance. Autocorrelation adjustment was not applied, consistent with the use of aggregated annual mortality data from CDC WONDER. Trends were considered increasing or decreasing if the slope of the regression differed significantly from zero, determined by 2-tailed t testing. Statistical significance was set at p < 0.05.

Reliability and validity of the study

The reliability and validity of this study are grounded in the rigor and universality of the database employed as well as the statistical methods utilized.CDC WONDER provides population-wide mortality data with standardized collection, duplicate verification, and uniform coding [26], ensuring reproducibility. JoinPoint Regression is a validated tool for epidemiological trend analysis, with consistent APC/AAPC estimation confirmed in prior CDC WONDER-derived cardiovascular studies [27, 28]. CMR/AAMR are established indicators for population-level mortality burden, consistent with aortic disease epidemiology standards [29]. Stratified measures align with known AA/AD epidemiological patterns.

All stratified analyses across age, sex, race/ethnicity, census region, and urban-rural status were exploratory and hypothesis-generating, and no formal multiplicity correction was applied given the descriptive nature of this population-based mortality trend study.

Results​

Annual trends​

A total of 289,971 adult deaths (≥ 25 years) attributed to AA and AD were recorded between 1999 and 2023. From 1999 to 2023, the AAMR showed a sustained downward trend, decreasing from 8.89 per 100,000 population in 1999 to 3.75 per 100,000 population in 2023 [AAPC: -3.73 (95% CI: -4.33, -3.14); p < 0.001]. A sharp decline was observed between 2005 and 2011 [APC: -7.65 (95% CI: -9.47, -5.80); p < 0.05], and the AAMR stabilized after 2021 (Fig. 1)(Table S1).

Fig. 1.

AAMR related to AA and AD per 100,000 population in the United States, overall and stratified by sex, 1999–2023; Both: Joinpoint 1 (2005)***, Joinpoint 2 (2011)***; Male: Joinpoint 1 (2003)***, Joinpoint 2 (2012)***; Female: Joinpoint 1 (2006)***, Joinpoint 2 (2011)***, representing the time points of significant changes in AA/AD mortality trends. p*<0.05, p**<0.01, p***<0.001

Sex-stratified AA- and AD-Related AAMR​

During the period from 1999 to 2023, the number of male deaths decreased from 9,592 to 5,978, with a total of 171,929 male deaths recorded; for females, the number declined from 6,179 to 4,104, with a total of 118,042 female deaths. Notably, both genders experienced a decrease of over 30%. AAMR decreased in both males and females, with males consistently having a higher overall AAMR (7.83 per 100,000 population) than females (3.31 per 100,000 population). Male AAMR declined from 13.57 per 100,000 population in 1999 to 5.00 per 100,000 population in 2023 [AAPC: -4.23 (95% CI: -4.80, -3.66); p < 0.001], while female AAMR decreased from 5.73 per 100,000 population in 1999 to 2.70 per 100,000 population in 2023 [AAPC: -3.29 (95% CI: -3.99, -2.58); p < 0.001].

In males, AAMR slightly decreased from 13.57 per 100,000 population in 1999 to 11.74 per 100,000 population in 2003 [APC: -3.94 (95% CI: -6.31, -1.50); p < 0.01], followed by a sharp decline to 5.97 per 100,000 population in 2012 [APC: -7.38 (95% CI: -8.30, -6.46); p < 0.001], and then maintained a steady downward trend until 2023. In females, AAMR decreased from 1999 to 2006 [APC: -2.89 (95% CI: -4.01, -1.76); p < 0.001], reaching 4.68 per 100,000 population in 2001, followed by a sharp decline to 3.19 per 100,000 population in 2011 [APC: -7.75 (95% CI: -10.52, -4.88); p < 0.001], and then a slow downward trend to 2.70 per 100,000 population in 2023 [APC: -1.60 (95% CI: -2.22, -0.98); p < 0.001] (Fig. 1)(Table S2).

Age-stratified AA- and AD-related mortality rates

When stratified by age group. During the study period, the number of deaths among individuals aged 85 years and older due to relevant diseases decreased from 3,337 in 1999 to 2,337 in 2023, representing a decline of approximately 30%. In total, 70,620 people succumbed to these diseases over the study period. During the study period, CMR increased substantially with age, with the highest rate observed in the ≥ 85 years group (51.11 per 100,000 population [AAPC: −3.13 (95% CI: −4.12, − 2.14); p < 0.001]), followed by the 75–84 years (26.70 per 100,000 population [AAPC: −4.87 (95% CI: −5.47, − 4.27); p < 0.001]), 65–74 years (10.36 per 100,000 population [AAPC: −5.12 (95% CI: −5.60, − 4.65); p < 0.001]), and 55–64 years groups (3.79 per 100,000 population [AAPC: −2.80 (95% CI: −3.43, − 2.16); p < 0.001]). Lower CMRs were observed in younger age groups, including 45–54 years (1.65 per 100,000 population [AAPC: 0.38 (95% CI: −0.72, 1.49); p = 0.5]), 35–44 years (0.77 per 100,000 population [AAPC: 1.04 (95% CI: 0.63, 1.465); p < 0.001]), and 25–34 years (0.27 per 100,000 population [AAPC: 0.13 (95% CI: −0.36, 0.62); p = 0.58]).

The CMR in the 85 + years age group decreased from 80.33 per 100,000 population in 1999 to 37.72 per 100,000 population in 2023.The fastest decline was observed between 2006 (68.25 per 100,000 population) and 2009 (50.75 per 100,000 population) [APC: -9.67 (95% CI: -15.99, -2.90); p < 0.01]. In contrast, the 35–44 years age group accounted for a cumulative total of 8,231 AA/AD-related deaths over the study period, representing approximately 2.84% of all such deaths. The annual death count in this group reached a preliminary peak of 282 cases in 2009 and a record high of 409 cases in 2023 (possible fluctuation due to low counts). Notably, the CMR in this age group exhibited a slow upward trend, increasing from 0.73 per 100,000 population in 1999 to 0.92 per 100,000 population in 2023 [AAPC: 1.04 (95% CI: 0.63, 1.465); p < 0.001]. No statistically significant changes in AAPC were observed in the 45–54 or 25–34 years age groups.(Fig. 2)(Table S3).

Fig. 2.

CMR related to AA and AD per 100,000 population by age group in the United States, 1999–2023;25–34 years: No significant joinpoints were detected in the AA/AD mortality trend for the study period (1999–2023). 35–44 years: No significant joinpoints were detected in the AA/AD mortality trend for the study period (1999–2023).45–54 years: Joinpoint 1 (2003)***, Joinpoint 2 (2011)***; 55–64 years: Joinpoint 1 (2010); 65–74 years: Joinpoint 1 (2013)**; 75–84 years: Joinpoint 1 (2005)***, Joinpoint 2 (2012)***; 85+ years: Joinpoint 1 (2006)**, Joinpoint 2 (2009)***, Joinpoint 3 (2020), representing the time points of significant changes in AA/AD mortality trends. p*<0.05, p**<0.01, p***<0.001

Race-stratified AA- and AD-related AAMR

During the study period, the total number of deaths was 27,632 for NH Black, 241,037 for NH White, 11,705 for Hispanic, and 8,885 for NH Other. Notably, significant differences in AAMR were observed across different racial and ethnic groups. Throughout the study period, NH Black individuals exhibited the highest overall AAMR (5.32 per 100,000 population [AAPC: −2.47 (95% CI: −3.82, − 1.10); p < 0.001]), followed by NH White individuals (4.76 per 100,000 population [AAPC: −3.72 (95% CI: −4.36, − 3.08); p < 0.05]). The NH Other group had an AAMR of 3.88 per 100,000 population [AAPC: −3.82 (95% CI: −5.06, − 2.56); p < 0.001], while Hispanic individuals exhibited the lowest AAMR (3.15 per 100,000 population [AAPC: −3.86 (95% CI: −4.75, − 2.96); p < 0.001]).

The AAMR among NH Black individuals showed an overall declining pattern with clear phase-specific changes, including a modest decrease during 1999–2006 (APC: −1.54 [95% CI: −2.89 to − 0.17]; p < 0.05), a pronounced decline during 2006–2009 (APC: −13.81 [95% CI: −23.04 to − 3.48]; p < 0.05), and a subsequent non-significant downward trend during 2009–2023 (APC: −0.32 [95% CI: −0.84 to 0.21]; p = 0.216). The Hispanic population experienced a significant and sustained reduction in AAMR during 1999–2014 (APC: −5.65 [95% CI: −6.63 to − 4.66]; p < 0.001), followed by a non-significant change during 2014–2023 (APC: −0.80 [95% CI: −2.74 to 1.18]; p = 0.406).The NH Other group exhibited consistently significant declines across all periods, including 1999–2006 (APC: −2.79 [95% CI: −5.04 to − 0.48]; <0.05), 2006–2011 (APC: −8.84 [95% CI: −13.64 to − 3.77]; p p < 0.01), and 2011–2023 (APC: −2.25 [95% CI: −3.14 to − 1.35]; p < 0.001), while the NH White population demonstrated persistent and significant declines during 1999–2006 (APC: −4.27 [95% CI: −5.25 to − 3.29]; p < 0.001), 2006–2011 (APC: −7.75 [95% CI: −10.29 to − 5.14]; p < 0.001), and 2011–2023 (APC: −1.66 [95% CI: −2.23 to − 1.08];

p < 0.001).(Fig. 3)(Table S4).

Fig. 3.

AAMR related to AA and AD per 100,000 population by race in the United States, 1999–2023; NH White: Joinpoint 1 (2006)***, Joinpoint 2 (2011)***; NH Black: Joinpoint 1 (2006), Joinpoint 2 (2009)*; Hispanic: Joinpoint 1 (2014); NH Other: Joinpoint 1 (2006)**, Joinpoint 2 (2011)***, representing the time points of significant changes in AA/AD mortality trends . p*<0.05, p**<0.01, p***<0.001

Geographic region-stratified AA- and AD-related AAMR

Census regions

Between 1999 and 2023, the total number of deaths was 55,188 in the Northeast, 58,213 in the West, 75,174 in the Midwest, and 101,396 in the South, the Midwest region had the highest overall AAMR (5.89 per 100,000 population [AAPC: -3.42 (95% CI: -4.01, -2.83); p < 0.001]), followed by the South (5.33 per 100,000 population [AAPC: -3.64 (95% CI: -4.23, -3.05); p < 0.001]), Northeast (5.28 per 100,000 population [AAPC: -4.02 (95% CI: -4.76, -3.28); p < 0.001]), and West (5.15 per 100,000 population [AAPC: -3.51 (95% CI: -4.45,-2.56); p < 0.001]) regions.

Across census regions, AAMR generally exhibited phased downward trends between 1999 and 2023, with notable regional heterogeneity. In the Midwest, AAMR declined steadily during 1999–2006 (APC: −4.20 [95% CI: −5.10, − 3.29]; p < 0.001), followed by a more pronounced decrease during 2006–2011 (APC: −7.28 [95% CI: −9.63, − 4.86]; p < 0.001), and subsequently a slower but still significant decline during 2011–2023 (APC: −1.30 [95% CI: −1.82, − 0.78]; p < 0.001). The Northeast demonstrated a similar pattern, with a significant reduction during 1999–2006 (APC: −3.97 [95% CI: −5.08, − 2.85]; p < 0.001), an accelerated decline during 2006–2011 (APC: −8.41 [95% CI: −11.32, − 5.41]; p < 0.001), and a sustained but more moderate decrease during 2011–2023 (APC: −2.16 [95% CI: −2.83, − 1.49]; p < 0.001). In the South, AAMR showed a modest and marginally non-significant decline during 1999–2003 (APC: −2.53 [95% CI: −5.02, 0.03]; p = 0.053), followed by a marked and significant decrease during 2003–2012 (APC: −6.81 [95% CI: −7.76, − 5.86]; p < 0.001), and then a slower yet significant downward trend during 2012–2023 (APC: −1.38 [95% CI: −2.05, − 0.71]; p < 0.001). In contrast, the West experienced a significant decline during 1999–2006 (APC: −3.79 [95% CI: −4.62, − 2.95]; p < 0.001), a sharp reduction during 2006–2009 (APC: −10.77 [95% CI: −16.87, − 4.22]; pp < 0.01), and a continued significant decrease during 2009–2020 (APC: −2.82 [95% CI: −3.40, − 2.24]; p < 0.001), whereas no significant change was observed during 2020–2023 (APC: 2.31 [95% CI: −1.51, 6.27]; p = 0.218).

Notably, the overall AAMR differences among the South, Northeast, and West regions were minimal. During the study period, AAMR showed an overall downward trend across all regions. The fastest decline was observed in the Northeast region (decreasing from 8.57 per 100,000 population in 1999 to 3.46 per 100,000 population in 2023), while the slowest decline occurred in the Midwest region (decreasing from 10.11 per 100,000 population in 1999 to 4.55 per 100,000 population in 2023)(Fig. 4)(Table S5).

Fig. 4.

AAMR related to AA and AD per 100,000 population by U.S. Census Region, 1999–2023; Northeast: Joinpoint 1 (2006)***, Joinpoint 2 (2011)***; Midwest: Joinpoint 1 (2006)***, Joinpoint 2 (2011)***; South: Joinpoint 1 (2003)***, Joinpoint 2 (2012)***; West: Joinpoint 1 (2006)**, Joinpoint 2 (2009)***, Joinpoint 3 (2020), representing the time points of significant changes in AA/AD mortality trends. p*<0.05, p**<0.01, p***<0.001

Urban-rural classification

Between 1999 and 2020, the total number of deaths was 52,591 in nonmetropolitan areas and 207,335 in metropolitan areas, the overall AAMR in Nonmetropolitan areas (6.32 per 100,000 population) was higher than that in Metropolitan areas (5.33 per 100,000 population). In Nonmetropolitan areas, the AAMR showed a fluctuating downward trend from 10.09 per 100,000 population in 1999 to 8.05 per 100,000 population in 2005 [APC: -3.70 (95% CI: -5.21, -2.17); p < 0.001], followed by a rapid decline to 5.03 per 100,000 population in 2012 [APC: -6.36 (95% CI: -8.07, -4.62); p < 0.001], and finally a slow decrease to 4.29 per 100,000 population in 2020 [APC: -1.61 (95% CI: -2.92, -0.27); p < 0.05]. In Metropolitan areas, the AAMR slowly decreased from 8.64 per 100,000 population in 1999 to 6.59 per 100,000 population in 2006 [APC: -3.90 (95% CI: -4.60, -3.20); p < 0.001], followed by a sharp decline to 4.90 per 100,000 population in 2009 [APC: -10.58 (95% CI: -15.86, -4.96); p < 0.01], and then a slow decrease to 3.44 per 100,000 population in 2020 [APC: -2.79 (95% CI: -3.23, -2.35); p < 0.001], after which it stabilized (Fig. 5)(Table S6).(note: Given that the CDC database only includes data up to 2020, we will limit our analysis to this time point.)

Fig. 5.

AAMR related to AA and AD per 100,000 population by urbanization level in the United States, 1999–2020; Nonmetropolitan: Joinpoint 1 (2005)***, Joinpoint 2 (2012)*; Metropolitanp: Joinpoint 1 (2006)**, Joinpoint 2 (2009)***, representing the time points of significant changes in AA/AD mortality trends. p*<0.05, p**<0.01, p***<0.001. (note: data for 2021–2023 were unavailable)

Discussion​

This retrospective analysis of mortality data from 1999 to 2023 identified key findings related to AA and AD-associated deaths. First, the overall AAMR exhibited a consistent downward trend, with the most rapid decline generally occurring between 2005 and 2011, followed by stabilization post-2021. Males consistently demonstrated a higher AAMR compared with females, with marked racial/ethnic disparities: NH-Black populations bore the greatest disease burden. Geographic disparities were also observed, with the Midwest region having the highest mortality rate. From 1999 to 2020, AAMR remained higher in Nonmetropolitan areas than in Metropolitan areas. Finally, age-stratified analysis revealed that the ≥ 85 years age group carried the highest disease burden, while the 35–44 years age group exhibited an upward trend in overall mortality (Table S7). Interpretation of subgroup-specific APCs should be cautious, as exploratory stratified analyses may carry a potential risk of type I error inflation without formal multiplicity correction.

Cigarette smoking and hypertension are the two most important modifiable risk factors for both AA and AD. Cigarette smoking accelerates elastic fiber degradation and vascular inflammation, while chronic elevation of blood pressure increases aortic wall stress and promotes medial degeneration, thereby playing a central role in AAA development and increasing the risk of AD [30–32]. Aging is strongly associated with aortic stiffening, atherosclerosis, and collagen remodeling, leading to sharply increased incidence and mortality of AA and AD [33, 34]. Male sex, dyslipidemia, and inherited connective tissue disorders further increase susceptibility [30, 35, 36]. Major complications include aneurysm rupture, thromboembolism, organ malperfusion, acute aortic regurgitation, stroke, and sudden death, all associated with high mortality [35].

Reducing the burden of AA and AD requires integrated preventive strategies. Strict blood pressure control is fundamental, as antihypertensive therapy lowers aortic wall stress and reduces disease progression and dissection risk [35]. Tobacco control is one of the most effective population-level interventions, with smoking cessation significantly decreasing AAA incidence and mortality [30]. Lifestyle modifications, including weight control, regular physical activity, and healthy diet, improve cardiometabolic profiles and reduce overall vascular risk; Targeted screening programs, particularly ultrasound screening for high-risk individuals, facilitate early detection and elective intervention, although current programs largely focus on older adults [36]. Pharmacologic therapy and timely surgical or endovascular repair remain essential for preventing catastrophic outcomes [35].

Between 1999 and 2023, the total number of AA and AD-related deaths increased, whereas the national AAMR exhibited an overall downward trend. First, this trend may be attributable to global tobacco control efforts, as smoking represents a modifiable risk factor [37]. Cigarette smoke has been shown to exacerbate atherosclerosis, leading to elastin fragmentation, aneurysm formation, rupture, and subsequent death [38]. Notably, tobacco control initiatives in the United States have been highly effective: the adult smoking rate decreased by 73%, from 42.6% in 1965 to 11.6% in 2022 [39]. Consistent with this, previous studies have demonstrated that the reduction in AA prevalence is concurrent with decreased cigarette and tobacco consumption, thereby reducing AA incidence at the population level [37]. Similarly, tobacco control measures have exerted a substantial protective effect on AD. Smoking may compromise aortic wall integrity by inducing endothelial dysfunction, exacerbating inflammatory responses, accelerating atherosclerosis, and elevating blood pressure, thereby increasing the risk of intimal tearing and dissection formation [31]. As adult smoking prevalence in the United States has declined markedly over recent decades, the burden of smoking-related vascular injury has decreased in parallel, which may contribute to a reduced population-level risk of AD and improved overall prognosis [40]. In addition, epidemiological evidence indicates that passive smoking is also associated with increased AD-related mortality, and that reductions in tobacco exposure are consistent with the long-term downward trend in AD mortality [41], supporting tobacco control as an important public health strategy for mitigating the burden of aortic dissection. Collectively, these findings indicate that sustained tobacco control efforts may have contributed substantially to the observed decline in age-adjusted mortality from both AA and AD among U.S. adults.

Second, advances in medical care and disease management have contributed, at least in part, to the observed decline in mortality. The introduction and widespread adoption of TEVAR have fundamentally transformed the treatment paradigm for type B aortic dissection and high-risk patients, leading to a substantial reduction in perioperative mortality [42, 43]. Concurrently, cross-sectional imaging modalities (computed tomography [CT], magnetic resonance imaging [MRI]) have become standard practice, enabling earlier and more accurate diagnosis even in atypical or incidental cases [44, 45]. These technological advancements have been complemented by the development of standardized care pathways and multidisciplinary “aortic teams,” which have optimized triage and intervention strategies [46]. Finally, the advancement of national cardiovascular disease screening and prevention policies in the United States has facilitated early detection and timely treatment, contributing to the observed trend.

Throughout the study period, significant gender disparities were observed, with males consistently maintaining a higher AAMR than females. First, this disparity is associated with the unique physiological characteristics of females. Estrogen has been shown to maintain aortic wall structural stability by inhibiting aortic smooth muscle cell apoptosis and reducing elastic fiber degradation (e.g., downregulating the activity of matrix metalloproteinase-9 [MMP-9]) [47]. On the other hand, risk exposure factors indicate that high-risk behaviors are more concentrated in males: first, the smoking rate among males (14.8% in males vs. 11.2% in females in 2020) is higher than that in females [48];second, males in the United States are more likely to have hypertension than females, while a higher proportion of females are aware of their hypertension diagnosis, have received treatment, and achieved hypertension control [49]. Additionally, females are generally more likely to seek medical care than males [50]. Finally, it is noteworthy that AAA are less common in females, they face a higher risk of aneurysm rupture and mortality than males with aneurysms of similar size [51].

In addition to AA, pronounced sex-based differences have also been observed in AD. Estrogen has protective effects on vascular biology, including anti-inflammatory actions, improved endothelial function, and reduced extracellular matrix degradation, which are believed to contribute to lower AD susceptibility in females [52]. Females generally present with AD at older ages than males, consistent with a later loss of estrogen’s vascular protection and later onset of hypertension and vascular stiffness. In contrast, male vascular biology and hemodynamic profiles—such as greater aortic wall shear stress and higher prevalence of uncontrolled hypertension—are associated with increased mechanical stress on the aortic wall and greater propensity for intimal tearing. These biological and hemodynamic differences, along with sex-specific risk factor profiles, likely contribute to the observed higher incidence of AD in men [53].

Our study revealed that age-stratified analysis aligned with the trend of higher mortality in older age groups alongside an overall downward trajectory. First, both AA and AD are typical age-related vascular diseases, with their incidence and mortality risks increasing exponentially with age [54]. This may be attributed to the high prevalence of comorbidities (e.g., hypertension, dyslipidemia, atherosclerosis) in this population, coupled with comprehensive physical function decline that precludes tolerance of surgical intervention. Furthermore, the 35–44 years age group exhibited an upward mortality trend, indicating an emerging younger-onset pattern of the disease. First, recent U.S. data demonstrate that the prevalence of hypertension in the 35–44 years age group increased from 12% in 2000 to 22% in 2023 [55]; long-term uncontrolled hypertension in young and middle-aged adults accelerates aortic wall injury, induces early-onset aortic dissection, and such patients exhibit faster progression of dissection rupture with higher mortality [56, 57]. Second, among U.S. adults aged 35–44 years, approximately 23.9% are current smokers, 27.4% engage in insufficient physical activity, and 44.4% report averaging ≤ 6 h of sleep per day [58]. Additionally, excessive alcohol consumption (e.g., binge drinking) is common in younger adults, affecting roughly 17–20% [59]. These lifestyle risk factors are strongly associated with increased cardiovascular and hypertension risk. Finally, existing aortic disease screening programs in the United States primarily target older populations, with nearly no age-specific screening available for the 35–44 years cohort [60]. The increasing prevalence of hypertension among younger adults and unhealthy lifestyle habits have emerged as potential contributors to the development and progression of AA and AD.

Additionally, stimulant use, including cocaine and amphetamines, acutely elevates blood pressure and has been linked to higher AD risk [61, 62]. National Survey on Drug Use and Health (NSDUH) data indicate that adults aged 35–44 have among the highest prevalence of illicit stimulant use, with past-year central nervous system stimulant use reported at ~ 9.5% for ages 35–39 and ~ 8.8% for ages 40–44, highlighting this age group as a key population for stimulant exposure [63].

Therefore, promoting healthy lifestyles among the general population and lowering the age threshold for targeted aortic disease screening are crucial to reducing the overall AAMR.

Regarding the racial-level changes in AAMR. First, prior to 2017, the NH White population exhibited a significantly higher degree of aging compared with the NH Black population, with the proportion of individuals aged 80 + years in NH White being 1.8 times that in NH Black [64]. As both AA and AD are typical “age-related diseases” with exponentially increasing incidence and mortality risks with age [54], this demographic difference likely contributed to the higher disease burden in NH White than in NH Black during this period. Second, U.S. aortic disease screening programs (e.g., abdominal ultrasound screening) have primarily prioritized NH White-dominant areas with adequate healthcare resources, resulting in a significantly higher AA/AD diagnosis rate in NH White than in NH Black [65]. However, subsequent trends indicate that NH Black populations may not have benefited equally from existing medical advances. On one hand, the overall income level and health insurance coverage rate of NH Black individuals are significantly lower than those of NH White and Hispanic populations. Meanwhile, the outpatient visit rate for aortic aneurysms and ultrasound screening rate among Black individuals are significantly lower than those among White individuals [66]. Third, hypertension and smoking are particularly prevalent among NH Black individuals—both are major risk factors for the development and progression of AA/AD. This increases the risk of AA/AD onset and severity, elevates treatment complexity, and reduces access to surgical intervention, ultimately resulting in a slower decline in AAMR among NH Black than in NH White [67–69].

In addition to socioeconomic and healthcare disparities, emerging evidence suggests that biological susceptibility may also contribute to the higher burden of aortic diseases in Black populations. Black individuals have been shown to exhibit greater arterial stiffness, as measured by pulse wave velocity, compared with White individuals, indicating earlier vascular aging and mechanical vulnerability of the aortic wall [70]. Impaired endothelial function, including reduced nitric oxide-mediated vasodilation and increased arterial wave reflections, has been observed in Black adults independent of traditional risk factors, which may further compromise vascular resilience [71]. Moreover, hypertension tends to develop earlier and with greater severity in Black populations, thereby prolonging exposure to elevated aortic wall stress that promotes medial degeneration and dissection risk [72]. Although genetic determinants of extracellular matrix remodeling and smooth muscle cell function may also modulate aortic wall integrity, these mechanisms remain underexplored across racial groups and warrant further investigation.

On the other hand, the slow decline in AAMR among NH Black individuals is also closely associated with the failure of their comorbidity control rate to improve synchronously with advances in medical technology. Therefore, targeted public health interventions and policies are urgently needed to improve access to prevention, diagnosis, and treatment services in underserved communities. This could include expanding affordable healthcare coverage, implementing community health education programs (e.g., hypertension management, smoking cessation), enhancing screening for high-risk ethnic groups, and promoting healthcare team diversity and culturally competent services to eliminate health disparities and achieve equitable health outcomes.

Finally, our study revealed the highest AAMR in the Midwest region geographically, with such disparities potentially driven by risk exposure, healthcare infrastructure and demographic structure. CDC BRFSS data indicate that adult obesity prevalence is highest in the Midwest and South, at approximately 36.0% and 34.7%, respectively, compared with 28.6% in the Northeast and 29.1% in the West [73]. According to 2016 BRFSS data, obesity prevalence in nonmetropolitan counties was 34.2%, significantly higher than the 28.7% observed in metropolitan counties, a pattern consistent across multiple Census regions [74]. Additionally, 2022 NHIS data estimate that approximately 19.8% of U.S. adults are current smokers; although smoking prevalence has declined overall, it remains relatively high in the Midwest and rural areas [75]; Concurrently, some areas within the Midwest exhibit a higher degree of aging, and older populations have reduced vascular elasticity and a higher incidence of atherosclerosis, which further elevates the baseline disease mortality rate [76]. Finally, compared with the Midwest, several states in the Northeast and South exhibit a higher concentration of healthcare resources, greater density of tertiary hospitals, and more developed emergency care networks for aortic diseases, which may contribute to the observed reduction in mortality [77].

Furthermore, AAMR was higher in Nonmetropolitan areas than in Metropolitan areas, reflecting prevalent socioeconomic challenges in Nonmetropolitan regions—such as higher poverty rates and lower educational attainment—which may limit access to healthcare and resources required for effective disease management, further contributing to elevated mortality [8, 78]. Additionally, Metropolitan areas typically have better health insurance coverage and financial capacity to afford advanced treatments, whereas such therapeutic options are less accessible in Nonmetropolitan areas—this discrepancy may exacerbate disparities in disease outcomes [79]. Further efforts to narrow the healthcare disparity between Nonmetropolitan and Metropolitan areas could be achieved by enhancing awareness among healthcare providers and patients, expanding training for surgeons specializing in endovascular techniques, and implementing policies such as subsidies to reduce barriers in underserved regions.

The geographic distribution of these behavioral risk factors aligns with regions of high chronic disease and AA/AD burden, suggesting that lifestyle behaviors and disparities in healthcare resources may partially explain regional differences in mortality.

Limitations

This study is subject to several limitations. First, the CDC WONDER database lacks individual-level risk factor and care variables (smoking history, BP control, screening/medication use, insurance, treatment type, hospital capability), precluding robust causal explanation of mortality trends and disparities. The database also lacks detailed individual socioeconomic status, precluding comprehensive confounding adjustment and accurate assessment of individual-level mortality drivers. Potential survivor bias exists in the 85 + age group, as this population represents a more health-resilient subset of the general population; individuals with severe AA/AD or multiple comorbidities may not survive to this advanced age, potentially underestimating the true AA/AD mortality burden in this group.

Second, despite the application of age adjustment, this study did not fully quantify the relative contributions of advancing diagnostic and therapeutic technologies to the observed mortality trends over the 25-year study period. The widespread clinical adoption of CT/MRI in the later years likely enhanced the identification of AA/AD cases, which may have led to undercounting in the early years (1999–2000 s) and thus created the illusion of a more pronounced mortality decline than the actual reduction in disease burden. Concomitantly, therapeutic advances—including the development and clinical use of TEVAR/EVAR, stent technologies, and multidisciplinary aortic care teams—have directly reduced fatal AA/AD outcomes. Notably, the CDC WONDER database lacks granular data on diagnostic modality utilization, treatment patterns, and hospital technological access, precluding the disentanglement of these artifactual and genuine effects on mortality. Additionally, temporal changes in healthcare utilization patterns throughout the study period were not fully accounted for in our analytical model. Additionally, while consistent ICD-10 coding was used throughout the study, gradual improvements in coding practices and physician awareness of AA/AD over 25 years may have enhanced case detection and documentation in later years, which could slightly attenuate the observed downward mortality trend.

Thirdly, as a population-based mortality database, mortality information is collected by state registries and provided to the National Vital Statistics System. Data are based on death certificates for U.S. residents.But CDC WONDER does not provide individual-level details on the specific diagnosing provider (e.g., name, years of experience) or the exact diagnostic workflow for each case.we cannot verify the individual experience level of each provider or the specific imaging/autopsy protocols used for each case.

Fourthly, a notable inherent limitation of the CDC WONDER database is the temporal inconsistency in urban-rural data: urban-rural mortality data were only available for the period 1999–2020, whereas data for other stratifications (sex, age, race/ethnicity, and region) covered 1999–2023, as the urban-rural classification data for 2021–2023 have not yet been released. This temporal discrepancy has somewhat reduced the comprehensiveness of the analysis on urban-rural disparities in recent years.

Finally, reliance on death certificate data from CDC WONDER is prone to misclassification bias of AA/AD-related deaths. We also cannot distinguish between incident and prevalent AA/AD cases, as the database contains no information on the timing of disease diagnosis, limiting interpretation of mortality trends related to disease onset versus progression. On one hand, while we restricted our analysis to AA/AD as the underlying cause of death to minimize misclassification, residual bias in death certificate coding of primary vs. secondary etiologies may still exist; on the other hand, although cases were identified using the ICD-10 code I71 series, variations in diagnostic coding practices and reporting standards across different regions may have impacted the accuracy of case inclusion and the comparability of mortality data.​.

Conclusion

From 1999 to 2023, overall AAMR for AA and AD steadily declined, with the greatest reduction observed between 2003 and 2012 and stabilization after 2021. Disease burden remained high among non-Hispanic Black individuals, males, and adults aged ≥ 85 years, with the Midwest exhibiting higher AAMR than other regions. Additionally, between 1999 and 2020, nonmetropolitan areas consistently showed higher burden than metropolitan areas. These findings highlight the need for targeted interventions, optimized disease management, and equitable healthcare to reduce disparities in AA/AD burden.​

Abbreviations

AAA

Abdominal Aortic Aneurysm

AAD

Acute Aortic Dissection

AAMR

Age-Adjusted Mortality Rate

APC

Annual Percentage Change

AA

Aortic Aneurysm

AD

Aortic Dissection

AAPC

Average Annual Percentage Change

CDC

Centers for Disease Control and Prevention

CI

Confidence Interval

CMR

Crude Mortality Rate

ICD-10

International Classification of Diseases, 10th Revision

IRB

Institutional Review Board

MMP-9

Matrix Metalloproteinase-9

NCHS

National Center for Health Statistics

NH Black

Non-Hispanic Black

NH Other

Non-Hispanic Other

NH White

Non-Hispanic White

TAA

Thoracic Aortic Aneurysm

TEVAR

Thoracic Endovascular Aortic Repair

WONDER

Wide-Ranging Online Data for Epidemiologic Research

Authors’ contributions

LL contributed to conceptualization, data curation, and formal analysis; XD G to methodology and investigation; JL to data curation and formal analysis; HL W to data curation; HJ Z to validation and visualization; SC Q to conceptualization, supervision, and funding acquisition; and HY L to conceptualization, supervision, and methodology, with all authors contributing to writing - review & editing.

Data availability

All data analyzed during this study were retrieved from the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research (CDC WONDER) database (https://wonder.cdc.gov/mcd.html). Further inquiries can be directed to the corresponding author.

Declarations

Ethical approval and consent to participate

No ethical approval was required for the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.