Aortic aneurysms in a general population cohort: prevalence and risk factors in men and women

Abstract

Aims

The prevalence and difference in risk factors for having thoracic aortic aneurysm (TAA) and abdominal aortic aneurysm (AAA) in men compared with women in the general population is not well described. This study aimed to test the hypotheses that (i) cardiovascular risk factors for TAA and AAA differ and (ii) the prevalence of TAA and AAA is sex specific.

Methods and results

Aortic examination using computed tomography angiography was performed in 11 294 individuals (56% women), with a mean age of 62 (range 40–95) years participating in the Copenhagen General Population Study. TAAs were defined as an ascending aortic diameter ≥45 mm and a descending aortic diameter ≥35 mm, while AAAs were defined as an abdominal aortic diameter ≥30 mm. Demographic data were obtained from questionnaires. Overall prevalence of aortic aneurysms (AAs) in the study population included: total population 2.1%, men 4.0% and women 0.7% (P-value men vs. women P < 0.001). AAs were independently associated with male sex, increasing age, and body surface area (BSA). While TAAs were associated with hypertension, odds ratio (OR) = 2.0 [95% confidence interval (CI): 1.5–2.8], AAAs were associated with hypercholesterolaemia and smoking, OR = 2.4 (95% CI: 1.6–3.6) and 3.2 (95% CI: 1.9–5.4).

Conclusion

Subclinical AAs are four times more prevalent in men than in women. In both sexes, increasing age and BSA are risk factors for AAs of any anatomical location. Whereas arterial hypertension is a risk factor for TAAs, hypercholesterolaemia and smoking are risk factors for AAAs.

Structured Graphical Abstract

Risk factors for and prevalence of thoracic aortic aneurysm and abdominal aortic aneurysm (AAA). The width of risk factor bars at each aortic location represents the odds ratio for ascending aortic aneurysm (AA), descending AA, and AAA. The right side of the illustration depicts the different AA locations and the prevalence stratified by sex.

Open in new tabDownload slide

Introduction

The incidence of cardiovascular disease (CVD) has been decreasing over the past two decades, whereas the prevalence is increasing and CVD remains the most common cause of death worldwide.¹ One important type of CVD, aortic aneurysms (AAs), may remain undetected and can lead to severe clinical manifestations resulting in high mortality rates. These clinical manifestations include aortic dissection and AA rupture with an ensuing mortality of up to 85% even after acute intervention.² It has been reported that abdominal aortic aneurysms (AAAs) account for 1.3% of all deaths in men aged between 65 and 85 years in developed countries.³ Thoracic aortic aneurysm (TAA) and AAA are asymptomatic in approximately 80–90% of patients, earning it the surname ‘silent killer’ due to its potential for catastrophic complications.^4,5 In cases where symptoms do present, TAA may include pain in the jaw, neck, chest, and upper back,⁶ whereas AAA may include pain in the abdomen, flank, groin, and legs.² However, risk factors may differ when comparing the pathogenesis in TAA and AAA⁷ as reflected by the histopathological differences in the aortic wall components when comparing thoracic and abdominal aorta.⁸ As the pace of population ageing in Western countries is much faster than ever before, the clinical burden from AAs is expected to increase.

AA is defined as a focal dilatation of the aorta greater than 150% of the expected normal aortic size^9–11. AAs may develop at any level of the aorta and are commonly divided into TAA and AAA. Although TAA and AAA are abnormalities found in the same vessel, they appear to have different pathologies and aetiologies.¹² AAs are believed mostly to develop in men; however, women with AAs are much worse than men,^13–15 yet the sex-specific prevalence and the relative importance of cardiovascular risk factors for development of TAA vs. AAA remain largely unresolved.

In the Copenhagen General Population Study (CGPS), we tested the hypotheses that: (i) the prevalence of TAA and AAA is sex specific and (ii) cardiovascular risk factors for TAA and AAA differ.

Methods

Study population

The CGPS is a cohort study with since 2003 more than 100 000 individuals enrolled from the Capital Region of Denmark. From 2010 onwards, CGPS participants were invited to a have a research computed tomography (CT) angiography (CTA) examination at Copenhagen University Hospital, Rigshospitalet, Denmark.¹⁶ Aortic analyses were performed in participants included between 2010 and 2019. A regional ethics committee (H-KF-01-144/01) and a local institutional review board approved the research protocol, and all individuals provided oral and written informed consent. Inclusion criteria to participate in the CT examination were age 40 years or above and normal kidney function (serum creatinine <100 µmol/L). Individuals that were ineligible for contrast administration during examination with only non-contrast CT images available were excluded from the study population. Participants with bicuspid aortic valve or previous aortic repair were excluded from this study.

Cardiovascular risk factors

At inclusion in the CGPS, participants completed detailed questionnaires on cardiovascular risk profile and prescribed medication, and body metrics were measured as previously described.¹⁶ Self-reported questionnaire questions with prescribed medication included whether the participant was taking anti-hypertensive, anti-diabetic, or cholesterol-lowering medication. Blood samples were drawn for biochemical tests, and body surface area (BSA) was calculated using the DuBois–DuBois formula. Hypertension was defined as systolic/diastolic blood pressure ≥ 140/90 mmHg¹⁷ measured on two separate days or prescribed anti-hypertensive medication. Hypercholesterolaemia was defined as low-density lipoprotein > 3.0 mmol/L or prescribed statins.¹⁸ Smoking status was defined as either being a current or former smoker. Diabetes was defined as self-reported diabetes or having a non-fasting plasma glucose ≥ 11.1 mmol/L.

CT protocol and image analysis

In the present study, contrast-enhanced CT images were used for image analysis of the entire aorta. Cardiac CTA scans were acquired during the period from 2010 to 2019. From 2014 onwards, the scan protocol was extended with an abdominal CTA. CT angiography image acquisition was performed using a 320-slice multidetector CT (Aquilion ONE, ViSION Edition, Canon Medical Systems, Japan) with the following scanner settings: gantry rotation time 275 ms and detector collimation 0.5 × 320. Choice of tube voltage (100–135 kV) and tube current (40–900 mA) was based on body mass index and body metrics on the scout images. Intravenous contrast media (Visipaque 78 mL, 320 mg/mL) was infused with a flow rate of 5 mL/s, followed by saline. A cardiac CTA was scanned electrocardiogram (ECG)-gated in the 75% phase of the R–R interval and reconstructed with 0.5 mm slice thickness and increments of 0.25 mm. An abdominal CTA was acquired using helical scan mode and reconstructed with 0.5 mm slice thickness and increments of 0.30 mm. Oral beta-blocker was administered to individuals with a heart rate > 60 bpm, and all received sublingual nitroglycerine prior to scanning.

Aortic measurements and identification of AAs were performed using imaging analysis software on dedicated workstations (Vitrea, v. 6.9, Vital Images Inc., MN, USA). CT images were analysed by three experienced readers. Blinded inter- and intraobserver variability of the aortic measurements have previously been reported.¹⁹ Aortic measurements were performed at maximum diameter in each aortic region. Ascending aorta was measured from sinotubular junction till aortic arch if visible or top of scan field, descending aorta from aortic arch if visible or top of scan field till diaphragm, and abdominal aorta from diaphragm segment of abdominal aorta till iliac bifurcation. Thoracic aortic measurements were performed on cardiac CTA images, which will be referred to as thoracic CTA, whereas abdominal aortic measurements were performed on the abdominal CTA images. TAAs were defined as an ascending aortic diameter ≥ 45 mm and a descending aortic diameter ≥ 35 mm, whereas AAAs were an abdominal aortic diameter ≥ 30 mm according to contemporary guidelines.^10,11,20 AA type was classified as fusiform or saccular.²¹ AA prevalence, location, distribution, combinations of concomitant AAs, and aneurysm type are reported stratified by sex.

Statistical analysis

Statistical analyses were performed using R version 4.1.3. Continuous variables with normal distribution were presented as mean (SD) and compared using Student’s t-test. Continuous variables with non-normal distribution were presented as median [inter-quartile range (IQR)] and compared using the Mann–Whitney U test. Categorical variables were presented as n (%) and compared using the χ² test. AA prevalence was calculated as (n TAA)/(n thoracic CTA) or (n AAA)/(n abdominal CTA). Whereas for the male-/female-specific prevalence, it was calculated as (n TAA in men)/(n thoracic CTA in men), which is similar for women, for abdominal aorta, it was calculated as (n AAA in in men)/(n abdominal CTA in men), which is similar for women. Logistic regression models were performed to obtain odds ratios (ORs) of each explanatory variable with AAs, followed by multivariable logistic regression models to obtain adjusted odds ratios (aORs) of multiple explanatory variables with AAs. Explanatory variables included sex, age, BSA, hypertension, hypercholesterolaemia, smoking, and diabetes. A multinomial logistic regression model was performed to test for covariates coefficient difference between ascending AA, descending AA, and AAA. P-values for difference in multinomial logistic regression coefficients were compared in ascending AA, descending AA, and AAA. Supplementary analyses using sex-stratified logistic regression models were performed where adjustments for age, BSA, hypertension, hypercholesterolaemia, smoking, and diabetes were included in a model for men and women. Aortic size index was calculated as an aortic diameter divided by BSA. These were used in a supplementary logistic regression analysis for AAs. Two-tailed P-values of <0.05 were considered statistically significant.

Results

A total of 12 895 eligible individuals from the CGPS underwent CT examination. Out of them, 1601 individuals (12%) were excluded due to unavailable contrast-enhanced CT, resulting in a total of 11 294 individuals in the study population (see Supplementary data online, Figure S1). All 11 294 individuals had a thoracic CTA, of whom 7442 individuals (66%) also had an abdominal CTA.

Clinical characteristics of the study population are presented in Table 1. The study population consisted of 56% women, and the median age (IQR) was 62 (53–69) years. Regarding cardiovascular risk factors, 36% of the study population had hypertension, 51% were current or former smokers, and 66% had hypercholesterolaemia. When comparing the clinical characteristics of the study population stratified by sex, men were taller, weighed more, and had higher blood pressure compared with women. Furthermore, the proportion of hypertension, hypercholesterolaemia, smokers, and diabetes were higher in men than in women as presented in Supplementary data online, TableS1.

Prevalence of AAs

In total, we found 354 individuals (3.1%) with AAs of any anatomical location. This included 273 (2.4%) TAAs (210 men and 63 women, P < 0.001) and 122 (1.6%) AAAs (109 men and 13 women, P < 0.001). The aneurysm prevalence, aortic location, combinations of AAs, and aneurysm type stratified by sex are presented in Table 2 and in the Structured Graphical Abstract. Men had a higher prevalence of AAs in all aortic locations compared with women. In men, AAAs (3.5%) were the most frequent aneurysm location followed by ascending AAs (3.0%) and descending AAs (1.2%). In women, ascending AAs (0.9%) were the most frequent aneurysm location followed by AAAs (0.3%) and descending AAs (0.1%). When comparing AA location distribution between men and women, P-values for difference were non-significantly different (P = 0.063). Clinical characteristics of individuals with and without ascending AA, descending AA, and AAA are given in Supplementary data online, TablesS2–S4.

Risk factors associated with AAs

Univariable ORs for AAs of each risk factor and anatomical location are presented in Table 3—left columns. Male sex [OR: 3.5 (2.6–4.8)], age per 10 years [OR: 1.8 (1.5–2.0)], BSA per 0.1 m² [OR: 1.4 (1.3–1.5)], hypertension [OR: 3.0 (2.3–4.0)], and hypercholesterolaemia [OR: 1.6 (1.1–2.2)] were associated with an increased risk of ascending AAs. But for descending AAs, male sex [OR: 11.4 (5.2–24.9)], age per 10 years [OR: 2.6 (2.0–3.3)], BSA per 0.1 m² [OR: 1.5 (1.4–1.7)], hypertension [OR: 3.8 (2.3–6.3)], and smoking [OR: 2.9 (1.6–4.9)] were associated with an increased risk of descending AAs. For AAAs, male sex [OR: 12.1 (6.8–21.5)], age per 10 years [OR: 3.5 (2.8–4.3)], BSA per 0.1 m² [OR: 1.5 (1.4–1.6)], hypertension [OR: 2.6 (1.8–3.8)], hypercholesterolaemia [OR: 4.1 (2.8–6.0)], and smoking [OR: 3.7 (2.4–5.7)] were associated with an increased risk of AAAs. All other covariates were not associated.

In fully adjusted multivariable logistic regression models, the aOR for AAs in each aortic location is presented in Table 3—right columns and Figure 1. The aORs showed that male sex, increasing age, and BSA remained significantly associated with increased OR for AAs in all aortic locations. Hypertension [ascending OR: 2.0 (1.5–2.8) and descending OR: 2.0 (1.2–3.4)] was associated with an increased risk of ascending and descending AAs, whereas hypercholesterolaemia [OR: 2.4 (1.6–3.6)] and smoking [OR: 3.2 (1.9–5.4)] were associated with an increased risk of AAAs. Diabetes [OR: 0.1 (0.02–0.9)] was associated with a decreased risk of AAAs.

Uni- and multivariable models for TAA and AAA in men and women separately are shown in Supplementary data online, TablesS5 and S6 and presented as forest plots in Supplementary data online, FiguresS2 and S3. P-values for interaction between sex and risk factors in TAA and AAA were all non-significantly different as presented in Table 4.

In Table 4, the results for interaction test of each risk factor associated with either ascending AA, descending AA, or AAA are presented with no AAs as reference group. Sex, age, hypertension, hypercholesterolaemia, and smoking showed coefficient difference between ascending AA and AAA, whereas coefficients for BSA and diabetes showed no difference in ascending AA compared with AAA.

Aortic size index

Supplemental analyses including aortic size index were performed. Aortic size index was different in men compared with women, as presented in Supplementary data online, Table S7. In additional supplemental statistical analysis, when including aortic size index in regression models, the results showed that increasing aortic size index was associated with an increased risk of AAs at all aortic locations, as presented in Supplementary data online, Table S8. When adjusting for aortic size index, data suggest that men are at higher risk of AAs and the risk is probably not related to larger body size.

Discussion

In this cross-sectional cohort study, 11 294 individuals from the CGPS were examined. TAA and AAA were more prevalent in men than in women. While male sex, increasing age, and BSA were common risk factors for both TAA and AAA, hypertension was associated with TAAs alone and hypercholesterolaemia, smoking, and diabetes (inverse) were associated with AAAs alone. When delving deeper into the AA characteristics stratified by sex, it is noteworthy that there was no interaction between sex and risk factors associated with AAs; thus, the risk factors for AAs were not sex specific.

Risk factors for regional AAs seem to be slightly different when comparing men and women. In women, hypertension was associated with ascending AAs, whereas for descending AA and AAA, only a borderline significant trend was observed. The prevalence of TAAs found in our study is comparable to previously reported prevalence of TAA in general population studies ranging from 1.2 to 1.4%.^22,23 The prevalence of AAAs was also comparable with a previous report of global and regional prevalence of AAAs ranging from 0.9 to 1.9%.²⁴

This study provides valuable new insights into the complex landscape of cardiovascular risk factors associated with AAs. In supplemental analyses, adjustments for aortic size index were made. The results showed that men remained at higher risk of AAs; therefore, the increased risk in men is probably not related to larger body size alone. While TAA and AAA are both manifestations of aortic disease, our findings illuminate the nuanced ways in which these conditions interact with various risk factors. The differential impact of risk factors on TAA and AAA underscores the necessity for a stratified approach to both research and clinical practice. Our results suggest that the pathophysiological mechanisms underlying TAA and AAA may differ more significantly than previously recognized.⁷ The differentiation in how specific cardiovascular risk factors contribute to each type of aneurysm challenges the traditional view that TAA and AAA are simply variations of the same disease. This could contribute to a paradigm shift in how medical professionals understand, screen, diagnose, and treat AAs.

Most previous studies on AAs were conducted only in elderly men and focused on either TAA or AAA separately. Our study included both men and women aged 40–95 years from the general population and evaluated both thoracic and abdominal aortic locations. Our study has several novel findings. We found (i) a higher prevalence of AAs at all aortic locations in men compared with women in the general population and (ii) a difference in the risk factor estimates for TAA vs. AAA.

Inflammatory and degenerative processes of the vascular tissue are part of the cellular mechanisms responsible for TAA and AAA formation.¹² Although TAA and AAA are two different locations of the same vessel, TAA and AAA may have different pathologies and aetiologies.²⁵ The microstructure of the aortic wall has been studied with great interest,^8,26–28 because alteration of the composition of connective fibers—particularly elastin and collagen—within the aortic wall directly correlates to the elasticity and strength of the aortic wall. There are three major aspects regarding the composition of elastin and collagen in the medial layer of the aortic wall: (i) the thickness and amount of elastin in the medial layer gradually decrease from proximal to distal aorta, resulting in a decrease of elastin-to-collagen ratio from the aortic root to the iliac bifurcation;²⁶ (ii) the amount of elastin decreases with ageing and in aneurysm formation;^8,27 and (iii) the collagen concentration also decreases with ageing, especially in abdominal aorta.²⁸ Therefore, the combination of alteration in microstructure of the aortic wall and vascular ageing may result in the abdominal aorta being more prone to vascular degenerative processes and ultimately lead to aneurysm formation. Accordingly, our findings of the difference in risk profiles of TAA and AAA may reflect the alteration of microstructural composition in the aortic wall, suggesting that abdominal aorta may be more prone to atherosclerotic risk factors.

AA is a complex multifactorial disease in which lifestyle, genetics, and cardiovascular risk factors play a significant role.²⁹ But, commonly known risk factors for AAAs include male sex, increasing age, Caucasian ethnicity, family history of AAA, smoking, and atherosclerosis.^30,31 However, which cardiovascular phenotypes carry the highest risk of developing aneurysms in the thoracic aorta compared with the abdominal aorta remains unclear. Furthermore, cardiovascular imaging biomarkers such as calcification score on CT play a critical role in assessing the risk and progression of cardiovascular and aortic diseases.³²

The prevalence of AAs in men aged 65–74 years in our study was comparable to the reported prevalence of AAs in the DANCAVAS study.^20,33 In comparison, women aged 65–74 years had a five-fold lower prevalence for AAs when compared with men in the same age group. Our results revealed that AAs were progressively more frequent with ageing, a finding that is consistent with the understanding of AAs being a degenerative process of the vascular wall. Screening for AAAs in men ≥ 65 years is recommended in several guidelines^10,34,35 and has been proved to reduce mortality and be cost-beneficial.^36,37 This has led to more countries implementing national screenings programmes forAAA.³⁸ A recent study showed that 28% of men with a sub-aneurysmal abdominal aorta, 26–29 mm, at the age of 65 years, are estimated to develop a large AAA (>55 mm) within the next 15 years.³⁹ Although rupture rates for AAAs are almost four-fold higher in women than in men,⁴⁰ screening high-risk women for AAAs may still not be cost-beneficial due to the low prevalence.^41,42 This is probably caused by a higher aortic size index in women compared with men, as the results shown in our study.

We investigated traditional risk factors for AAs including male sex, increasing age, and BSA and found them to be associated with AAs as previously described.^33,43 A possible physiological explanation for this association is that greater body size is associated with larger arteries, which have greater circumferential wall stress as described by the law of Laplace.

Hypertension has previously been reported to be more strongly associated with TAAs than AAAs.²⁵ Accordingly, we found that hypertension was independently associated with only ascending AAs in fully adjusted models. It was also associated with AAAs in unadjusted models but revealed non-significant difference in fully adjusted models. The mechanism of how hypertension contributes to the development of TAAs is unknown. A previous study using magnetic resonance velocity mapping revealed that the degree of wall shear stress may differ considerably among different aortic locations.⁴⁴ The presence of hypertension may further impair the wall shear stress and exacerbate the weakness of the vessel wall, leading to the formation of AA.

The inverse association between diabetes mellitus and AAA has previously been reported.⁴⁵ We also found that diabetes was independently associated with a decreased risk of AAAs, suggesting that diabetes mellitus or diabetic medication may slow down the development of AAAs. The exact molecular and cellular effect of diabetes mellitus or diabetic medication in modulating the mechanisms of aneurysm formation remains unknown. It is suggested that diabetes mellitus itself contributes to vascular tissue remodelling and anti-inflammatory processes, whereas anti-diabetic drugs interfere with the pathophysiological mechanisms of aneurysm formation.⁴⁶ Future studies of clinical relevance could be to look at ECG-defined left ventricular hypertrophy and its association with AAs.

Strengths and limitations of the study

The study’s utilization of detailed demographic, biochemical, and CT data allowed for a multifaceted analysis of the population. A significant strength of this study is the large sample size, which offers robust statistical power. This increases the likelihood of detecting true associations and provides more precise estimates of effects. As this study was a single-centre study and only in Caucasian individuals, generalizability to other populations may be limited. Data on medication were based on self-reported questionnaire data; therefore, some participants may have incorrectly answered the questionnaire. Data on family history of aneurysms were unavailable, which is a limiting factor in regard to patients with heritable aortic conditions and could potentially bias the results. As thoracic aortic measurements were performed on cardiac CTA scan images, that did not cover the entire thoracic aorta, and therefore, a limitation to potential missed TAAs in the regions not visible on the scan images. Data on systemic inflammatory conditions and aortitis were unavailable and could potentially affect the results as these non-genetic factors play a role in aneurysm formation.

Clinical perspectives

Understanding that risk factors may have varying degrees of influence on TAA and AAA formation could inform more targeted screening programmes. Such programmes could consider the individual’s risk profile, including their sex and other personal risk factors, potentially leading to earlier detection and more successful management of AAs. The new evidence provided by this study can contribute to refining existing risk stratification models for AAs. By integrating the differential effects of risk factors on TAA and AAA, clinicians could more accurately predict an individual’s risk and tailor management accordingly. Furthermore, although the prevalence of AAs was lower in women compared with men when using a uniform threshold for AAs, the AA location distribution and risk factors associated with AAs were indifferent in women compared with men, and this may suggest that sex-specific AA thresholds are needed for earlier detection of aortic disease in especially women. This study warrants further aortic research and studies to elucidate the underlying mechanisms in TAA and AAA as well as potential clinical applications of individualized aortic risk stratification in aortic screening programmes.

Conclusion

In a general population study, TAA and AAA were more prevalent in men compared with women. Male sex, increasing age, and BSA were common risk factors for both TAA and AAA. Hypertension was independently associated with TAAs, whereas hypercholesterolaemia, smoking, and diabetes (inverse) were independently associated with AAAs.

Author contributions

Conception and design, analysis and interpretation of the data, drafting of the article, critical revision of the article for important intellectual content, and final approval of the article: M.H.C.P., P.E.S., A.F., J.T.K., H.S., S.A., B.G.N., L.V.K., and K.F.K.; provision of study materials or patients and administrative, technical, or logistic support: M.H.C.P., P.E.S., A.F., J.T.K., S.A., B.G.N., and K.F.K.; statistical expertise and collection and assembly of data: M.H.C.P., P.E.S., S.A., B.G.N., and K.F.K.; and obtaining of funding: M.H.C.P., P.E.S., and K.F.K.

Supplementary data

Supplementary data are available at European Heart Journal - Cardiovascular Imaging online.

Funding

This study was funded by AP Møller og Hustru Chastine McKinney Møllers Fond and Rigshospitalet.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

References

Author notes