Journal of Vascular Surgery
Volume 51, Issue 1 , Pages 3-7, January 2010

Genetic and environmental contributions to abdominal aortic aneurysm development in a twin population

  • Carl Magnus Wahlgren, MD, PhD

      Affiliations

    • Department of Vascular Surgery, Karolinska University Hospital, Stockholm, Sweden
    • Corresponding Author InformationReprint requests: Carl Magnus Wahlgren, MD, PhD, Department of Vascular Surgery, Karolinska University Hospital, 171 76 Stockholm, Sweden
    • ,
  • Emma Larsson, MD

      Affiliations

    • Department of Vascular Surgery, Karolinska University Hospital, Stockholm, Sweden
    • ,
  • Patrik K.E. Magnusson, PhD

      Affiliations

    • Department of Medical Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden
    • ,
  • Rebecka Hultgren, MD, PhD

      Affiliations

    • Department of Vascular Surgery, Karolinska University Hospital, Stockholm, Sweden
    • ,
  • Jesper Swedenborg, MD, PhD

      Affiliations

    • Department of Vascular Surgery, Karolinska University Hospital, Stockholm, Sweden

Received 2 June 2009; accepted 15 August 2009. published online 25 November 2009.

Article Outline

Background

The contribution of hereditary and environmental factors to the development of abdominal aortic aneurysms (AAAs) is still partly unknown. The aim of this study was to analyze the role of these factors in a large population-based sample of twins.

Methods

The Swedish Twin Registry, containing data on twins born in the country since 1886, was cross-linked with the Inpatient Registry, providing national coverage of discharge diagnoses coded according to the International Classification of Diseases (ICD). All twins with an infrarenal AAA were identified. Concordance rates and tetrachoric correlations were calculated for monozygotic (MZ) and dizygotic (DZ) twins. Tetrachoric correlations were calculated assuming an underlying normal distribution of liability, with multiple factors contributing additively and a threshold value that discriminates between AAA and no AAA. Higher concordance rates and correlations of liability in MZ twins than in DZ twins suggest that genetic factors influence disease development. Structural equation modeling techniques, Mx-analyses, were used to estimate the contributions of genetic effects as well as shared and nonshared environmental factors for development of AAA.

Results

There were 172,890 twins registered at the time of the study including 265 twins (81% men; mean age 72 years; range, 48-94) with AAA. There were 7 MZ and 5 DZ concordant pairs as well as 44 MZ and 197 DZ discordant pairs with AAA. The probandwise concordance rates for MZ and DZ pairs were 24% and 4.8%, respectively. The tetrachoric correlations were 0.71 in MZ pairs and 0.31 in DZ pairs. The odds ratio (OR) was 71 (95% confidence interval [CI] 27-183) for MZ twins and 7.6 (95% CI 3.0-19) for DZ twins. In the structural equation models, genetic effects accounted for 70% (95% CI 0.33-0.83), shared environmental effects for 0% (95% CI 0-0.27), and nonshared environmental effects for 30% (95% CI 0.17-0.46) of the phenotypic variance among twins.

Conclusion

These data provide robust epidemiologic evidence that heritability contributes to aneurysm formation. Concordances and correlations were higher in MZ compared with DZ twins, indicating genetic effects. There was a 24% probability that an MZ twin of a person with AAA will have the disease. The twin of an MZ twin with AAA had a risk of AAA that was 71 times that of the MZ twin of a person without AAA. A heritability of 70% of the total trait variance was estimated. The remaining variance was explained by nonshared environmental factors with no support for a role of shared environmental influences.

 

Abdominal aortic aneurysm (AAA) affects approximately 5% of elderly men and is estimated to be the tenth most common cause of mortality in the Western world.1 The prevalence of AAA found in population-based ultrasound scan screening studies ranges from 4.2% to 8.8% in men and 0.6% to 1.4% in women.2, 3 Risk factors associated with AAA include age, atherosclerosis, smoking, and family history of AAA.4 Familial clustering of patients with AAA supporting an inheritable disease has been well described.5, 6 It has been estimated that around 15% of patients with AAA have a family history of AAA disease.7 Evaluation of familial occurrence of AAA from reports mainly based on single-center experience has shown that siblings of patients with AAA, especially among brothers, are more affected.6 A population-based case control study including more than 3000 cases and 15,000 controls, found an approximately two-fold increased risk for AAA among first-degree relatives.8 The gender of the index person or the first-degree relative did not influence the risk for AAA in this study.

The mechanism underlying aneurysm formation will likely involve interplay between genetic predisposition and environmental risk factors.9 Risk factors analyzed in population AAA screening studies such as hypertension, coronary heart disease, and diabetes are partly explained by genetic factors.1, 10, 11 Smoking, dietary habits (high cholesterol levels), sporadic mutations, and infections are examples of risk factors often considered to be environmental, but might also be influenced by genes.10, 12 The contribution of hereditary and environmental factors to the development of AAA is still unclear. The aim of this study was to analyze the role of these factors in a large population-based sample of twins.

Back to Article Outline

Patients and methods 

Data sources and study design 

The Swedish Twin Registry (STR) was established in the late 1950s and contains data on almost all twins born in Sweden between 1886 and 1990.13 The registry has been expanded and updated on several occasions. The classical twin method is based on the fact that monozygotic (MZ) twins share 100% of their genome, whereas dizygotic (DZ) twins share, on average, 50% of their segregating genes.14 Twin similarity in the STR was determined by asking twins whether they and their twin partner were similar as “two peas in a pod” during childhood or not more than siblings in general. If both twins answered “similar as two peas in a pod,” they were classified as MZ. Twin pairs in which both twins answered that they were not more similar than siblings in general were classified as DZ. Those who answered differently were classified as not determined zygosity. Methods for assigning zygosity have been validated in the registry several times with an accuracy of 95% to 99%.13 Analyses of the relative importance of genes and environment for a phenotype can be carried out with traditional quantitative methods that are well developed for twin studies (Table I).

Table I. Glossary for statistical twin analysis
Intra-class correlationA statistical measure for strength and direction of resemblance between two variables (twin pair). It can vary between −1 and +1.10, 16
Structural equation modeling (SEM)A method that estimates regression coefficients (“parameters”) between latent (unobserved) and observed variables. These estimates minimize the difference between the covariance structure of the observed data and that predicted by the model.10, 16
Hereditary effectsThe proportion of phenotypic variance accounted for by inherited genetic differences among persons.12, 16
Shared environmental effectsThe proportion of phenotypic variance accounted for by environmental factors shared by both twins, thus contributing to similarity between them (eg, similar dietary habits [high cholesterol] or passive smoking in childhood family).12, 16
Non-shared environmental effectsThe proportion of phenotypic variance accounted for by environmental factors causing differences between twins (eg, sporadic mutations, infections, smoking, or occupational exposure).12, 16

Modified from References 10, 12, 16.

The National Board of Health and Welfare's Hospital Discharge Register (National Patient Register [NPR]) contains data on individual hospital discharges and covers all public in-patient care in the country. The register now includes 50 million discharges (for the period 1964 to 2006).15 In 2003, the main diagnosis was missing in 0.9% of the hospital admissions reported. The diagnoses coded according to the International Classification of Diseases (ICD)-7 until 1968, ICD-8 between 1969 and 1986, ICD-9 between 1987 and 1996, and ICD-10 from 1997 until now were used. The NPR is continuously validated by the National Board of Health and Welfare, and reporting is mandatory. An AAA was coded as 451 in ICD-7, 441.20 in ICD-8, 441D and 441E in ICD-9, and I71.3 and I71.4 in ICD-10.

The two registries (STR and NPR) were cross-linked using the national 10-digit social security number, a unique number assigned to every Swedish citizen. The study was approved by the local research ethics committee and by the STR Board.

Prevalence 

The prevalence of AAA was calculated as a percentage of individual twins diagnosed in the entire twin cohort as well as for the MZ and DZ group of twins separately.

Probandwise concordance 

Twin similarity was calculated as probandwise concordance rates for both zygosity groups. Concordance rates represent the probability of developing AAA for an individual with an affected twin. Concordant pairs are pairs in which both twins experienced the event under study. Discordant pairs are pairs in which one twin experienced the event under study.

When MZ concordance rates are greater than DZ concordance rates, genetic influences are indicated. If the concordance is similar for both types of twins and higher than expected by chance, then environmental effects are indicated.

Odds ratio 

The relative risk of AAA for persons whose twin had AAA compared with persons whose twin did not was estimated as an odds ratio (OR) and was calculated as:

in which a is the number of concordant pairs, b and c are each half the number of discordant pairs, and d is the number of concordant pairs without AAA.

Intraclass correlation and heritability 

Tetrachoric correlations were calculated assuming an underlying normal distribution of liability, with multiple factors contributing additively, and a threshold value that discriminates between AAA and no AAA. Differences in correlations between various groups provide information about the presence of genetic effects. Higher correlations of liability in MZ twins than in DZ twins suggest that genetic factors influence disease development.

Structural equation models can determine the variation of genetic and environmental influences in a phenotype.16 According to standard biometric practice; no gene-gene interaction, no gene-environment interaction or correlation, random mating, and the assumption that twins are representative of the general population were all assumed.16 A biometric path model is shown in the Fig. The total variance in the trait can be partioned into genetic variance (a2), shared (familial) environmental variance (c2), and nonshared (unique) environmental variance (e2).13 The method for selecting the best model followed standard procedures by using the Mx structural equation analyses.10 In order to estimate the parameters of interest, the equation for one of the twins can be written as:

  • View full-size image.
  • Fig. 

    Biometric path model. P, Phenotype; G, genetic variance; C, shared environmental variance; E, nonshared environmental variance; MZ, monozygotic; DZ, dizygotic; TRA, twins reared apart; TRT, twins reared together.

Vp is the total phenotypic variance of the population, representing the sum of the individual components a2, c2, and e2. Heritability, the relative importance of genetic influences for variation in a trait, is defined as genetic variance (a2) divided by the total phenotypic variance.

We could estimate the contribution of genetic (A) as well as shared (C) and nonshared (E) environmental effects by using the full biometric model (ACE) and its submodel (AE). Because the effects of genetic dominance and shared environment are confounded in the classical study of twins reared together, it is not possible to estimate those parameters simultaneously in a single model.17 Shared environmental factors contribute to similarity in pairs of twins (eg, passive smoking in childhood family or similar dietary habits). Nonshared environmental factors are, for example, sporadic mutations, infections, or occupational exposure.

Statistical software 

Structural equation modeling was carried out with Mx-analyses, a software package specifically designed for analysis of genetically informative data.

Back to Article Outline

Results 

We identified 172,890 twins (51% women) at the time of the study including 277 twins with AAA (prevalence in the entire twin cohort: 0.16%; in the MZ twin group 1.2%; in the DZ twin group 1.4%). Twelve twins had unknown zygosity and were excluded from the study. The final study sample included in the analysis was 265 twins with AAA: 81% men; mean age 72 years (range, 48 to 94-years-old). There was no difference between MZ and DZ twins regarding the age of AAA diagnosis (P = .86). There were approximately 26% ruptured AAAs based on ICD-9 and 10. Among concordant pairs, there were 7 MZ (86% men) and 5 DZ (90% men; 1 pair opposite sex) with AAA. Corresponding numbers for discordant pairs were 44 MZ (77% men) and 197 DZ (80% men; 100 pairs opposite sex).

Twin similarity and correlations of liability for AAA are shown in Table II. The probandwise concordance rates for MZ and DZ pairs were 24% and 4.8%, respectively. The tetrachoric correlations were 0.71 in MZ pairs and 0.31 in DZ pairs. The increased concordance rates, as well as higher correlations of liability in MZ compared to DZ twins, provide strong evidence of a genetic influence in disease expression. The OR was 71.1 (95% CI 27.5-183.4) for MZ twins and 7.6 (95% CI 3.0-19.2) for DZ twins.

Table II. Twin similarity and correlations of liability for abdominal aortic aneurysm
ZygosityConcordant pairs with no AAAConcordant pairs with AAADiscordant pairs with AAAProbandwise concordanceTetrachoric correlation (95% CI)
Monozygotic49137440.240.71(0.55-0.86)
Dizygotic14,82051970.0480.31(0.15-0.48)

AAA, Abdominal aortic aneurysm; CI, confidence interval.

When analyzing twins >55 years of age, there were 6 MZ and 5 DZ concordant pairs with AAA. Corresponding numbers for discordant pairs were 23 MZ and 188 DZ. The OR was 36.2 (95% CI 13.3-98.3) for MZ twins and 4.5 (95% CI 1.8-11.4) for DZ twins.

The results of structural equation model-fitting to the data are shown in Table III. The heritability (the proportion of the variance attributable to genetic effects) for AAA was 70% in both models. The remaining variation among twins could be attributed to nonshared environmental effects (30%), but no effect of familial (shared) environmental factors.

Table III. Estimation of genetic and environmental effects for abdominal aortic aneurysm
ModelA: Genetic effects (95% CI)C: Shared environmental effects (95% CI)E: Nonshared environmental effects (95% CI)
ACE model0.70(0.33-0.83)0(0-0.27)0.30(0.17-0.46)
AE model0.70(0.54-0.83)—0.30(0.17-0.46)

CI, Confidence interval.

Back to Article Outline

Discussion 

The classical twin method is based on the fact that MZ twins share identical genotypes, whereas DZ twins are no more genetically alike than normal siblings. Furthermore, the use of statistical models to analyze data from large samples of twins makes it possible to estimate the magnitude of the genetic and environmental effects on susceptibility to AAA.12 In the largest AAA population-based twin study to date, we could provide robust epidemiologic evidence that heritability contributes to aneurysm formation. Concordances and correlations were significantly higher in MZ compared with DZ twins, indicating genetic effects. There was a 24% probability that the MZ twin of a person with an AAA will have the disease. The twin of an MZ twin with an AAA had a risk of an AAA that was 71 times that of the MZ twin of a person without an AAA. When looking at twins over the age of 55 with an AAA, then possibly excluding genetic connective tissue disorders such as Ehlers-Danlos and Marfan syndrome, the OR was still significantly higher for MZ twins compared to DZ twins. A heritability of 70% of the total trait variance was estimated. The remaining variance was explained by nonshared environmental factors with no support for a role of shared environmental influences.

AAAs are frequently familial.5, 6, 18 Clifton was the first to report that AAA could be an inheritable disease, describing three brothers with ruptured aneurysms.19 There is an increased incidence of AAA observed in first-degree relatives of affected individuals.20 When comparing the family histories of patients with AAA and controls, 19% of patients with AAA reported to have a first-degree relative with a history of AAA compared to only 2.4% of the controls.21 Ultrasound scan screening of siblings of patients with AAA shows an overall aneurysm incidence of 19% to 25%.22, 23 We have previously shown in a population-based study that 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, which is in agreement with findings for first-degree relatives in screening studies.1, 2, 8 The possible genetic predisposition to AAA was investigated by Powell and Greenhalgh,24 recording the family history from 60 consecutive patients presenting for aneurysm repair. The younger age of presentation and the concentration of women amongst those with a positive family history argued for multifactorial inheritance of AAA, suggesting the genetic component to contribute to 70% of the overall susceptibility.

In the literature, we could only find 6 MZ twins pairs with AAA reported.25, 26, 27, 28 The hereditary pattern of 50 families with two or more first-order relatives affected with AAA, including three pairs of identical twins, was studied by Tilson and Seashore.25 One of the twin pairs also had an affected mother. They suggested that only one gene is responsible for AAA inheritance, which is likely to be autosomal dominant. In addition, case reports of MZ twins with AAA have been reported.26, 27, 28

The twin model provides powerful means of examining the total genetic contribution to a given disease, especially a complex trait such as AAA.10, 29 Unlike family studies or the study of sibling pairs, potential confounders such as the variability of disease prevalence with age are removed.30 The twin model approach relies on important assumptions (random mating, no interaction between genes and environment, and equivalent environments for MZ and DZ twins) that potentially could over- or underestimate the genetic and environmental components.11 There was only one female concordant twin pair with AAA, probably explained by a high frequency of AAA in men in general and a relative small sample size in this study. We could, therefore, not draw any conclusions about a possible stronger heritability in men or women. Missing AAA diagnoses in the twin registry could possibly underestimate the genetic effects. It should be noted that our estimates are population-specific. In other regions, the proportions of the type of effects could differ because of different environmental factors. For the aneurysmal disease with several genetic and environmental factors, the liability model assumes that the disease will occur when there are enough contributory factors to push the individual's liability above the threshold.30

This study has analyzed the contribution of hereditary and environmental factors for AAA development. It emphasizes the importance of evaluating family history in patients with AAA and supports ultrasound scan screening of unaffected siblings. The data can improve the information given to first-degree relatives of patients with AAA regarding the risk of AAA. However, many questions in the genetics of familial AAA remain unanswered. The understanding how risk factors and genotype contribute to AAA development is important for patient selection to screening and more aggressive treatment.1, 7 Twin studies are a valuable source of information about the genetic basis of complex traits. In that way, phenotypes can be defined to maximize chances of successful gene mapping.

Back to Article Outline

Author contributions 


Conception and design: CW

Analysis and interpretation: CW, PM

Data collection: PM, CW

Writing the article: CW

Critical revision of the article: CW, EL, RH, JS, PM

Final approval of the article: CW, EL, RH, PM, JS

Statistical analysis: PM, CW

Obtained funding: Not applicable

Overall responsibility: CW

Back to Article Outline

References 

  1. Golledge J, Muller J, Daugherty A, Norman P. Abdominal aortic aneurysm: pathogenesis and implications for management. Arterioscler Thromb Vasc Biol. 2006;26:2605–2613
  2. Ashton HA, Buxton MJ, Day NE, Kim LG, Marteau TM, Scott RA, et al. The Multicentre Aneurysm Screen Study (MASS) into the effect of abdominal aortic aneurysm screen on mortality in men: a reandomised controlled trial. Lancet. 2002;360:1531–1539
  3. Fleming C, Whitlock EP, Beil TL, Lederle FA. Screening for abdominal aortic aneurysm: a best-evidence systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2005;142:203–211
  4. Lederle FA, Johnson GR, Wilson SE, Chute EP, Littooy FN, Bandyk D, et al. Prevalence and associations of abdominal aortic aneurysm detected through screening (Aneurysm Detection and Management (ADAM) Veterans Affairs Cooperative Study Group). Ann Intern Med. 1997;126:441–449
  5. van Vlijmen-van Keulen CJ, Pals G, Rauwerda JA. Familial abdominal aortic aneurysm: a systematic review of a genetic background. Eur J Vasc Endovasc Surg. 2002;24:105–116
  6. Kuivaniemi H, Shibamura H, Arthur C, Berguer R, Cole CW, Juvonen T, et al. Familial abdominal aortic aneurysms: collection of 233 multiplex families. J Vasc Surg. 2003;37:340–345
  7. Thompson AR, Drenos F, Hafez H, Humphries SE. Candidate gene association studies in abdominal aortic aneurysm disease: a review and meta-analysis. Eur J Vasc Endovasc Surg. 2008;35:19–30
  8. Larsson E, Granath F, Swedenborg J, Hultgren R. A population-based case-control study of the familial risk of abdominal aortic aneurysm. J Vasc Surg. 2009;49:47–50discussion 51
  9. Sandford RM, Bown MJ, London NJ, Sayers RD. The genetic basis of abdominal aortic aneurysms: a review. Eur J Vasc Endovasc Surg. 2007;33:381–390
  10. Boomsma D, Busjahn A, Peltonen L. Classical twin studies and beyond. Nat Rev Genet. 2002;3:872–882
  11. Zdravkovic S, Wienke A, Pedersen NL, de Faire U. Genetic influences on angina pectoris and its impact on coronary heart disease. Eur J Hum Genet. 2007;15:872–877
  12. Lichtenstein P, Holm NV, Verkasalo PK, Iliadou A, Kaprio J, Koskenvuo M, et al. Environmental and heritable factors in the causation of cancer–analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med. 2000;343:78–85
  13. Lichtenstein P, De Faire U, Floderus B, Svartengren M, Svedberg P, Pedersen NL. The Swedish Twin Registry: a unique resource for clinical, epidemiological and genetic studies. J Intern Med. 2002;252:184–205
  14. Martin N, Boomsma D, Machin G. A twin-pronged attack on complex traits. Nat Genet. 1997;17:387–392
  15. The National Patient Registry, The Centre for Epidemiology, The National Board of Health and Welfare, Stockholm, Sweden. www.socialstyrelsen.se
  16. Neale MC, Cardon LR. Methodology for genetic studies of twins and families. Dordrecht, The Netherlands: Kluwer Academic Publishers BV; 1992;
  17. Larsen TB, Sørensen HT, Skytthe A, Johnsen SP, Vaupel JW, Christensen K. Major genetic susceptibility for venous thromboembolism in men: a study of Danish twins. Epidemiology. 2003;14:328–332
  18. Norrgård O, Rais O, Angquist KA. Familial occurrence of abdominal aortic aneurysms. Surgery. 1984;95:650–656
  19. Clifton MA. Familial abdominal aortic aneurysms. Br J Surg. 1977;64:765–766
  20. Salo JA, Soisalon-Soininen S, Bondestam S, Mattila PS. Familial occurrence of abdominal aortic aneurysm. Ann Intern Med. 1999;130:637–642
  21. Johansen K, Koepsell T. Familial tendency for abdominal aortic aneurysms. JAMA. 1986;256:1934–1936
  22. Frydman G, Walker PJ, Summers K, West M, Xu D, Lightfoot T, et al. The value of screening in siblings of patients with abdominal aortic aneurysm. Eur J Vasc Endovasc Surg. 2003;26:396–400
  23. Baird PA, Sadovnick AD, Yee IM, Cole CW, Cole L. Sibling risks of abdominal aortic aneurysm. Lancet. 1995;346:601–604
  24. Powell JT, Greenhalgh RM. Multifactorial inheritance of abdominal aortic aneurysm. Eur J Vasc Surg. 1987;1:29–31
  25. Tilson MD, Seashore MR. Fifty families with abdominal aortic aneurysms in two or more first-order relatives. Am J Surg. 1984;147:551–553
  26. Borkett-Jones HJ, Stewart G, Chilvers AS. Abdominal aortic aneurysms in identical twins. J R Soc Med. 1988;81:471–472
  27. Inzaghi A, Montemezzani M, Zappa M, Cantoni G, Marconato R, Longo T. Atherosclerotic aortic-bisiliac aneurysms in two monocorial twins. Int Angiol. 1990;9:292–294
  28. Ricci MA, Nutter B, McDonald I, Strindberg G. Abdominal aortic aneurysm with left iliac aneurysm in identical twins. Ann N Y Acad Sci. 1996;800:254–255
  29. Luft FC. Twins in cardiovascular genetic research. Hypertension. 2001;37(2 Part 2):350–356
  30. Millard TP, Bataille V, Snieder H, Spector TD, McGregor JM. The heritability of polymorphic light eruption. J Invest Dermatol. 2000;115:467–470

 Competition of interest: none.

 The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a competition of interest.

PII: S0741-5214(09)01678-4

doi:10.1016/j.jvs.2009.08.036

Journal of Vascular Surgery
Volume 51, Issue 1 , Pages 3-7, January 2010

Article Tools

* Image Usage & Resolution