A population-based case-control study of the familial risk of abdominal aortic aneurysm
Article Outline
Background
Several studies have reported a familial clustering of abdominal aortic aneurysm (AAA) supporting that AAA is an inheritable disease, but few population-based studies can be found. Possible gender differences regarding hereditary patterns have been reported.
Objective
The aim of this study was to investigate the risk of developing an AAA for first-degree relatives of patients with AAA in Sweden and compare them with matched controls and their relatives.
Methods and Materials
All persons (3183) born after 1932, diagnosed with AAA between 2001 and 2005, and a random selection of 15,943 age-, gender-, and region-matched controls were included. First-degree relatives of cases and controls were identified via the Multigeneration Register. Family history of AAA for cases and controls was assessed by linking the relatives to the Hospital Discharge Register and Cause of Death Register. The data were analyzed by conditional logistic regression.
Results
The overall relative risk of AAA associated with family history compared to no family history was 1.9 (95% confidence interval [CI] 1.6-2.2). Comorbidities were more common among the cases than the controls (P < .0001) but the relative risks remained unchanged after adjustment for comorbidities. Stratification for absence or presence of comorbidities showed no significant difference between the two groups (P = .29). The relative risk of AAA for first-degree relatives was similar for women and men (P = .22 for gender differences), ie, the relative risk of AAA was not dependent on the gender of the index person.
Conclusion
In this nationwide survey, the relative risk of developing AAA for first-degree relatives to persons diagnosed with AAA was approximately doubled compared to persons with no family history. Neither the gender of the index person nor the first-degree relative influenced the risk of AAA.
Familial clustering of abdominal aortic aneurysm (AAA) was described more than 30 years ago.1 Case series have reported increased familial risk for AAA and several studies have reported a high prevalence of AAA among siblings of AAA patients.2, 3, 4 Case-control studies have shown an approximately four-fold increased risk being affected for first-degree relatives of AAA patients.5, 6 These studies are mostly based on consecutive patient series from one or few centers, and few population-based reports covering the subject can be found.
Possible gender differences regarding hereditary patterns have been reported, but the results from these studies are probably not possible to generalize.7, 8 Other important gender differences are known for AAA; men have a higher prevalence and are younger at onset of disease compared to women.9, 10
In order to perform a population-based case-control study, longitudinal data from nationwide registers is necessary. Sweden has a Multigeneration Register including first-degree relatives of persons born after 1931. This register, together with the Cause of Death Register and the In-patient Register, was utilized to perform a case-control study with the aim of investigating the risk to develop an AAA associated with a positive family history and possible gender differences.
Methods
National registries
All Swedish citizens have a unique 12-digit personal identification number, enabling identification of patients in national registries. The Swedish National Hospital Discharge Register, managed by the Swedish Board of Health and Welfare (NBHW), covers all information regarding public in-patient care. The Cause of Death Register is managed by the Centre for Epidemiology, NBHW, and covers data for cause and time of death for all deceased persons. The Multigeneration Register, managed by Statistics Sweden, contains linkages between children and parents and consists of persons in Sweden born after 1931 and their parents. The extraction of data was based on the diagnostic codes classified by the International Classification of Diseases (ICD) (Table I).
Table I. Diagnostic codes classified by the International Classification of Diseases (ICD)
| ICD | ||||
|---|---|---|---|---|
| 7 | 8 | 9 | 10 | |
| AAA | 451.00 | 441.20 | 441.3 | I71.3 |
| 441.4 | 171.4 | |||
| Aortic aneurysm⁎ | 452 | 442 | 441.5 | 171.8 |
| 441.6 | 171.9 | |||
| COPD | 500-502 | 490-492 | 491.2 | J44 |
| 491.8 | ||||
| 496 | ||||
| IHD | 420 | 410-414 | 410-414 | I20-I25 |
| Heart failure | 431.10 | 428.99 | 428 | I50 |
| Hypertension | 440-447 | 440-404 | 401-405 | I10-I15 |
⁎Without given localization. |
Cases (n = 3183) were defined as all persons found in the Swedish National Hospital Discharge Register with a first admission under the diagnosis of AAA or an aortic aneurysm without a given localization (I71.3, I71.4, I71.8, or I71.9) between 2001 and 2005, and all persons deceased during the same period with the corresponding underlying or contributing cause of death that not was found in the hospital register. A random selection of 15,943 age-, gender-, calendar year-, and region-matched controls were identified from the Register of Total Population. The Swedish population was 8.9 million (50.5% women) in 2001 and 9.0 million (50.4% women) in 2005.
ExposuresFirst-degree relatives of cases and controls were identified via the Multigeneration Register. Family history of aortic aneurysm for cases and controls was assessed by linking the relatives to the Hospital Discharge Register and Cause of Death Register (for codes used see Table I). Comorbidity for cases and controls was assessed by the prevalence of admissions with selected diagnoses 5 years prior to the diagnosis date for the case and the corresponding period for the matched controls (for codes used see Table I).
StatisticsCrude comparisons of proportions were performed by χ2 tests. Significance was defined as P < .05. The risk of AAA in relation to family history was estimated by means of conditional logistic regression and odds ratios (ORs) with corresponding 95% confidence intervals (CIs). Risks associated with family history were assessed with respect to number of affected relatives and the gender of affected relatives. In stratified analyses with respect to the subjects' age, gender, and diagnosis the potential effect modification of these factors were investigated and tested by likelihood ratio tests.
Risks in relation to family history were adjusted for presence of comorbidity. The categorization of the comorbidity according to the different comorbidities alone and combinations of two, three, or four comorbidities was chosen due to the lack of multiplicative effects of the different comorbidities. We also performed an interaction analysis of the effect of family history among cases and control with and without any comorbidity; this interaction was tested by a likelihood ratio test. The study was approved by the Regional Ethics Committee.
Results
Among the 3183 cases with AAA, 479 were identified in the Cause of Death Register and 2704 in the Swedish National Hospital Discharge Register. A total of 840 cases were diagnosed with ruptured AAA. Baseline characteristics are listed in Table II.
Table II. Characteristics for cases and controls
| Characteristics | Cases (n = 3183) | Controls (n = 15,943) | P |
|---|---|---|---|
| No (%) | No (%) | ||
| Gender | |||
 Females | 535(16.8) | 2679(16.8) | .9951 |
| Age | |||
| <40 | 29(0.9) | 145(0.9) | .999 |
| 41-49 | 54(1.7) | 269(1.7) | |
| 50-59 | 509(16.0) | 2531(15.9) | |
| 60-69 | 2031(63.8) | 10,196(64.0) | |
| 70-73 | 560(17.6) | 2802(17.6) | |
| Family history | |||
| Yes | 268(8.4) | 743(4.6) | <.0001 |
| Number of affected relatives | |||
| 0 | 2915(91.6) | 15,200(95.3) | <.0001 |
| 1 | 254(8.0) | 728(4.6) | |
| 2 | 12(0.4) | 14(0.1) | |
| 3 | 2(0.1) | 1(0.01) | |
| Comorbidity | |||
| No comorbid condition | 2215(69.6) | 14,179(88.9) | <.0001 |
| IHD | 297(9.3) | 652(4.1) | |
| Hypertension | 229(7.2) | 483(3.0) | |
| Heart failure | 50(1.6) | 88(0.6) | |
| COPD | 45(1.4) | 77(0.5) | |
| Any 2 comorbidities | 269(8.5) | 375(2.4) | |
| Any 3 comorbidities | 72(2.3) | 83(0.5) | |
| All 4 comorbidities | 6(0.2) | 6(0.04) | |
| Diagnosis | |||
| Rupture | 840 | ||
| With family history | 61(7.3) | ||
| Non-rupture | 2343 | ||
| With family history | 207(8.8) |
A positive family history for AAA was significantly more common among cases than controls (8.4 vs 4.6%, P < .0001). The cases had a significantly higher prevalence of comorbidities compared to the controls. The proportion of cases with family history was similar in patients with ruptured and non-ruptured AAA (Table II).
The overall relative risk of AAA associated with family history compared to no family history was approximately doubled (OR 1.9, 95% CI 1.6-2.2) (Table III). The risk increased further if more than one first-degree relative had AAA. The relative risks remained unchanged after adjustment for comorbidities (Table III).
Table III. Relative risk of AAA
| Odds ratio (95% CI) | ||
|---|---|---|
| Unadjusted | Adjusted for comorbidity | |
| Family history | ||
| No | 1.0(ref) | 1.0(ref) |
| Yes | 1.9(1.6-2.2) | 1.9(1.6-2.2) |
| Number of affected relatives | ||
| 0 | 1.0(ref) | 1.0(ref) |
| 1 | 1.8(1.6-2.1) | 1.8(1.6-2.1) |
| 2 | 4.7(2.1-10.2) | 4.3(1.9-9.8) |
| 3 | 10.0(0.9-110.3) | 12.2(1.1-137.0) |
In a logistic regression analysis stratified for absence or presence of comorbidities, there was no significant difference between the two groups, ie, the relative risk of AAA associated with family history was similar for cases suffering from comorbidities or not (P = .29) (Table IV).
Table IV. Relative risk of AAA associated with family history stratified for comorbidity or not
| Odds ratio (95% CI) | ||
|---|---|---|
| Number of affected relatives | Comorbidity = no | Comorbidity = yes |
| 0 | 1.0(ref) | 1.0(ref) |
| 1 | 1.7(1.5-2.1) | 2.4(1.7-3.3) |
| ≥2 | 4.8(2.0-11.5) | 4.2(0.8-21.8) |
In a subgroup analysis, no significant gender difference in the relative risk of AAA associated with family history could be demonstrated (P = .22 for differences between genders), ie, the relative risk of AAA was not dependent on the gender of the index person (Table V). No differences between age groups in the relative risk of AAA associated with family history were recorded (P = .75) (Table V). Patients treated for rupture and non-rupture had similar relative risks of positive family history of AAA (P = .42 for differences between rupture and non-rupture) (Table V).
Table V. Relative risk of AAA for subgroups given family history (ref: no family history in the specific subgroups)
| Odds ratio (95% CI) | ||
|---|---|---|
| Unadjusted | Adjusted for comorbidity | |
| Gender | ||
| Male | 1.8(1.5-2.1) | 1.9(1.6-2.2) |
| Female | 2.3(1.6-3.3) | 2.1(1.5-3.1) |
| Age | ||
| <49 | 2.7(1.0-7.3) | 2.4(0.8-7.1) |
| 50-64 | 1.8(1.5-2.3) | 1.9(1.5-2.3) |
| 65-73 | 1.9(1.5-2.3) | 1.9(1.5-2.3) |
| Diagnosis | ||
| Rupture | 1.7(1.3-2.4) | 1.8(1.3-2.4) |
| Non rupture | 1.9(1.6-2.3) | 1.9(1.6-2.3) |
There was a tendency towards a higher risk of being affected with AAA when having a male relative with AAA, compared to having a female relative (P = .07) (Table VI).
Table VI. Relative risk of AAA given a certain gender of the index person and the first-degree relative (ref: no family history)
| Odds ratio (95% CI) | ||
|---|---|---|
| Female relative | Male relative | |
| Male index person | 1.6(1.3-2.0) | 2.1(1.7-2.5) |
| Female index person | 1.8(1.1-2.9) | 3.1(1.9-5.0) |
Discussion
It is of major importance for families and the society to clarify whether first-degree relatives of AAA patients have an increased risk to develop AAA compared to the general population. We found that a positive family history of AAA was associated with an approximately doubled risk of AAA. Comorbidities and gender of the index person did not modify the relative risk associated with family history.
Familial tendency of AAA was first described by Clifton in 1977.1 Since then, several studies have reported a high prevalence of AAA among siblings of AAA patients, supporting that AAA is an inheritable disease.2, 3, 4 In a case-control study of 98 AAA cases and 102 controls, a positive family history of AAA was associated with an increased risk of AAA (OR 4.77, 95% CI 1.26-18.1).5 Similar risk rates were confirmed in a Finnish study of relatives of 150 AAA patients, but only 29% of the first-degree relatives were included in the study.6 In a Canadian report, an approximately four-fold higher prevalence of AAA was found in siblings of 126 AAA patients compared to siblings of a control group.11 The doubled risk of AAA associated with a positive family history in the present study confirms previously reported familial aggregation, but is lower than previously reported four-fold increase. Since AAA is a disease affecting the elderly, one concern with the present study is that the upper age limit of the included AAA cases was 73-years-old, but there were no age restrictions regarding the relatives. The present study covers the entire Swedish population; still misclassification is possible due to unknown relatives, eg, immigrants. The Swedish Multigeneration Register includes all persons born after 1931 and their parents; the coverage is 86% for known mothers and 83% for known fathers. The main effect of unknown relatives would be an underestimation of the effect of family history. There is also a risk of unreported cases with ruptured AAA due to low autopsy rates (14% in 2005) which is likely to add to the underestimation of the risk.
In a previous Swedish report by Hemminki regarding AAA from 1987 to 2001 using the same registers as in the present study, an increased standardized incidence ratio for siblings of AAA patients was found.12 Their study design only included 71 affected sib-pairs between 0 and 69-years-old and included all types of aortic aneurysm.
In the present study, the risk of AAA associated with family history was neither influenced by the gender of the index person nor the gender of the first-degree relative. This indicates that presence of family history should be taken into consideration regardless of gender when evaluating the risk to develop an AAA. It is, however, important to stress that our findings are based on index cases aged 73 years or younger. Since women develop AAA at an older age than men, a possible underestimation would be more pronounced for women. In two previous studies regarding the occurrence of AAA in relatives of AAA patients, the highest prevalence was found in brothers of female patients, but the differences were not significant.3, 13 A retrospective study on multiplex aneurysms without any control group reported a higher percentage of females among the affected relatives than among AAA probands.8 In a prospective case-control study, there was an increased prominence of women in the familial AAA group.7 However, the study group of 542 patients was from one center and the control group consisted of 500 spouses, non-blood relatives, and vascular patients without aneurysms which could have influenced the finding. Screening for AAA is cost effective for men above the age of 65, but this has not been shown to be the case for women. It is possible, and even likely, that the increased risk for women with a positive family history makes screening cost effective also for this group.
AAA is a multifactorial disease involving environmental and genetic risk factors. The present study design did not, however, address the genetic mechanisms. The increased risk for patients with more than one affected relative compared to only a single relative with AAA in our study supports that AAA is a multigenic disease. A recent study reported that a sequence variant on chromosome 9 (9p21) was associated with myocardial infarction, intracranial aneurysm, and AAA. The estimated degree of the effect differed and was most pronounced for AAA.14 Kuivaniemi et al15 performed a DNA linkage study for familial AAA in 233 families with more than one individual diagnosed with AAA. Linkage to chromosomes 19p13 and 4q31 were found, indicating that these regions could possess genetic risk factors for AAA. When investigating the manners of inheritance, they found that autosomal recessive inheritance was most common.
Patients with AAA often have manifestations of atherosclerotic disease, such as hypertension and coronary artery disease.9, 16 In the present study, AAA cases had more comorbidities which reflects that AAA patients often share the same risk factors as patients with other manifestations of cardiovascular disease. Since most AAAs are asymptomatic, the probability for detection will be influenced by contacts with healthcare providers. The occurrences of comorbidities obviously affect the frequency of such contacts, leading to an increasing number of diagnostic procedures. Regardless of the occurrence of comorbidities, the relative risk of AAA for first-degree relatives remained doubled in a stratified analysis. This emphasizes that family history is of major concern regardless if the person suffers from other cardiovascular diseases or not, as shown by others.5
In conclusion, in this nationwide survey the relative risk to develop AAA for first-degree relatives to persons diagnosed with AAA was doubled compared to persons with no family history. Neither the gender of the index person nor the first-degree relative influenced the risk of AAA.
Author contributions
References
- . Familial abdominal aortic aneurysms. Br J Surg. 1977;64:765–766
- The lifetime prevalence of abdominal aortic aneurysms among siblings of aneurysm patients is eightfold higher than among siblings of spouses: an analysis of 187 aneurysm families in Nova Scotia, Canada. J Vasc Surg. 2005;42:891–897
- The value of screening in siblings of patients with abdominal aortic aneurysms. Eur J Vasc Endovasc Surg. 2003;26:396–400
- . Selection for screening for familial aortic aneurysms. Br J Surg. 1992;79:897–898
- . Risk factors for abdominal aortic aneurysm: results of a case-control study. Am J Epidemiol. 2000;151:575–583
- . Familial occurrence of abdominal aortic aneurysm. Ann Intern Med. 1999;130:637–642
- Are familial abdominal aortic aneurysms different?. J Vasc Surg. 1989;10:39–43
- French women from multiplex abdominal aortic aneurysm families should be screened. Ann Surg. 2005;242:739–744
- Aneurysms of the abdominal aorta in older adults (The Rotterdam Study). Am J Epidemiol. 1995;142:1291–1299
- . Predictors of repair and effect of gender on treatment of ruptured abdominal aortic aneurysm. J Vasc Surg. 2004;39:784–787
- . Sibling risks of abdominal aortic aneurysm. Lancet. 1995;346:601–604
- . Familial risks of aortic aneurysms among siblings in a nationwide Swedish study. Genet Med. 2006;8:43–49
- The low prevalence of abdominal aortic aneurysm in relatives in Northern Ireland. Eur J Vasc Endovasc Surg. 2007;34:163–168
- The same sequence variant on 9p21 associates with myocardial infarction, abdominal aortic aneurysm, and intracranial aneurysm. Nat Genet. 2008;40:217–224
- Familial abdominal aortic aneurysms: collection of 233 multiplex families. J Vasc Surg. 2003;37:340–345
- . Smoking, atherosclerosis, and risk of abdominal aortic aneurysm. Eur Heart J. 1997;18:671–676
The study was funded by financial support provided by the Swedish Heart-Lung Foundation (Go Red), by the European Society of Vascular Surgery, and through the regional agreement on medical training and clinical research (ALF) between Stockholm County Council and the Karolinska Institutet.
Competition of interest: none.
PII: S0741-5214(08)01362-1
doi:10.1016/j.jvs.2008.08.012
© 2009 Published by Elsevier Inc.
