- Email: [email protected]
Plasma concentrations of apolipoproteins A-I, B and M in patients with abdominal aortic aneurysms
Clinical Biochemistry 43 (2010) 407–410
Contents lists available at ScienceDirect
Clinical Biochemistry j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c l i n b i o c h e m
Plasma concentrations of apolipoproteins A-I, B and M in patients with abdominal aortic aneurysms Josefin Ahnström a, Anders Gottsäter b, Bengt Lindblad b, Björn Dahlbäck a,⁎ a b
Wallenberg Laboratory, Department of Laboratory Medicine, Clinical Chemistry, Lund University, SE-205 02 Malmö, Sweden Vascular Centre Malmö-Lund, Malmö University Hospital UMAS, Malmö, Sweden
a r t i c l e
i n f o
Article history: Received 6 August 2009 Received in revised form 22 October 2009 Accepted 10 November 2009 Available online 22 November 2009 Keywords: ApoM Signal peptide HDL Apolipoprotein Lipoprotein Lipocalin AAA ApoA-I ApoB Aneurysmal disease
a b s t r a c t Objectives: Apolipoproteins play important roles in the development of atherosclerosis but their involvement in the pathogenesis of abdominal aortic aneurysm (AAA) is poorly understood. The aim was to investigate whether apoA-I, apoB and apoM are independently associated with AAA. Design and methods: Plasma apoA-I, apoB and apoM were measured in 343 patients with AAA and in 214 elderly apparently healthy control individuals from the background population. Results: AAA patients had lower apolipoprotein levels, as compared to healthy individuals: apoA-I, 1.62 vs. 2.08 g/L; apoB, 0.91 vs. 1.04 g/L; apoM, 0.72 vs. 0.91 μmol/L (p b 0.0001 for all three). In multivariate analyses, apoA-I and apoB were associated with AAA, odds ratios (95% confidence intervals) being 0.53 (0.43–0.64) and 0.86 (0.75–0.998), respectively. Conclusions: ApoA-I, apoB and apoM levels were significantly lower in patients with AAA than in the control individuals, but only apoA-I and apoB were independently associated to AAA. © 2009 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Abdominal aortic aneurysm (AAA) has increasingly been recognized as an important cause of mortality with an annual mortality in the US of around 15,000 individuals. The prevalence in an aged population (N60 years of age) is around 2–4% [1–4]. AAA usually occurs in the infrarenal part of the aorta and is generally defined as a maximal aortic diameter of ≥3 cm [2,4–6]. The underlying problem in aneurysmal disease is weakening of the aortic wall, resulting in progressive dilation and, if left untreated, risk for eventual aneurysm rupture. The principal determinant of AAA rupture is the maximum diameter of the aneurysm [7]. Male gender, smoking, hypertension, increasing age, coronary heart disease (CHD) and a family history of abdominal aortic aneurysm have been identified as independent risk factors for AAA in several population-based studies [1–5]. However, data on the association between lipid abnormalities and the risk of AAA are conflicting [2,4,5,8–11], even though it is well recognized that oxidized lowdensity lipoprotein (LDL) is a major cause of injury to vascular
⁎ Corresponding author. Wallenberg Laboratory, Division of Clinical Chemistry, Department of Laboratory Medicine, University Hospital, Lund University, Entrance 46, Floor 6, Malmö S-20502 Malmö, Sweden. Fax: +46 40 337044. E-mail address: [email protected] (B. Dahlbäck).
endothelium and smooth muscle cells in the media [12]. In CHD, apoA-I and apoB, the major apolipoproteins of high density lipoproteins (HDL) and LDL, respectively, have been shown to be equivalent or better than the HDL-cholesterol (HDL-c) and LDLcholesterol (LDL-c) as risk markers for atherosclerosis or cardiovascular events. ApoA-I and apoB have, as far as we know, only been measured in one study of AAA, which demonstrated decreased apoA-I and increased apoB levels in the patients [8]. The relatively recently discovered apolipoprotein M (apoM) is preferentially bound to HDL but it is also, to a minor extent, associated with the other lipoproteins [13,14]. ApoM is present in around 5% of HDL and in 2% of LDL particles [14]. The apoM concentration in human plasma is approximately 0.9 μmol/L (∼23 mg/L) with somewhat lower levels in young women than in older women and men [15]. It correlates positively to total plasma cholesterol and to the plasma concentrations of both LDL- and HDL-cholesterol [15,16]. In vivo studies in mice suggested apoM to play a role in HDL metabolism and to reduce the development of atherosclerosis [17,18]. Studies on isolated apoM-containing human HDL particles showed that those particles were more efficient in reducing LDL oxidation and stimulating cholesterol efflux than apoM-free HDL [14]. Despite the strong correlation of plasma apoM to total plasma cholesterol concentrations, and the positive results seen in mice, two independent case–control studies of CHD did not identify apoM as a predictor of disease [16].
0009-9120/$ – see front matter © 2009 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2009.11.006
408
J. Ahnström et al. / Clinical Biochemistry 43 (2010) 407–410
Aortic aneurysm disease shares many risk factors with cardiovascular diseases, but it is a distinct entity with unique pathogenic mechanisms. The aim of this study was to investigate whether there is a relationship between plasma apoA-I, apoB and apoM levels and the development of AAA. Materials and methods Study design AAA patients undergoing routine ultrasound follow-up at the Vascular Centre, Malmö University Hospital from February 2002 to December 2006 were included in the study. All investigations and blood sampling were performed prior to any invasive treatment. No patients with acute symptomatic or ruptured AAA were included. The study included 343 patients with either large AAA planned for operation or small AAA undergoing surveillance. The AAA patients were compared to individuals drawn from a background population selected from a large population study investigating preventive factors for cardiovascular disease and alcohol abuse [19]. The individuals, consisting of 214 elderly apparently healthy individuals without symptomatic AAA, cardiovascular disease or atherosclerosis, were included during 2004–2005 and were chosen to represent the general population of approximately the same age as the patients. Further data are given in Table 1. The Ethical Committee of Lund University approved the study and all patients gave written consent to participate. Measurements At baseline, all patients were evaluated regarding blood pressure (BP) (measured in the right arm after 10 min rest), diabetes mellitus (treatment of diabetes or a fasting blood glucose level of ≥6.7 mmol/L) and pharmacological treatment with cardiovascular drugs. Body mass index (BMI) was calculated as kg/m2. Smoking was defined as current smoking. Blood sampling was performed at baseline in all participants. Blood samples in appropriate tubes were directly handled and centrifuged (at 4 °C). Routine laboratory analyses were performed immediately. Samples for non-routine analyses were frozen in −80 °C for later analysis. Standard laboratory methods at the Department of Clinical Chemistry, Malmö University Hospital (SWEDAC approved according to European norm 45001), were used for analyses including
Table 1 Data for the subjects included in the study given as mean ± SD or proportion.
Age (years) Men (%) Diabetes mellitus (%) Smokers (%) Statin treatment (%) AAA-max (mm) BMI (kg/m2) Systolic BP (mmHg) Diastolic BP (mmHg) Cholesterol (mmol/L) HDL-c (mmol/L) Non-HDL-c (mmol/L) ApoA-I (g/L) ApoB (g/L) ApoM (μmol/L) IL-6 (pg/mL) TNF-α (pg/mL) CD40L (pg/mL) n.a., not applicable. a Mann–Whitney test. b χ2 test.
Cases (n = 343)
Controls (n = 214)
p-value
74 ± 8 79 11.9 34.6 38.7 62.8 ± 14.6 25.4 ± 4.05 141.1 ± 19.4 81.3 ± 10.5 4.98 ± 1.22 1.06 ± 0.35 3.95 ± 1.21 1.62 ± 0.36 0.91 ± 0.25 0.72 ± 0.21 9.36 ± 32.19 2.73 ± 6.56 582 ± 2229
68 ± 2 46.3 5.7 12.2 13.1 n.a. 27.1 ± 4.33 143.4 ± 19.7 84.9 ± 10.4 5.85 ± 1.07 1.24 ± 0.34 4.59 ± 1.02 2.08 ± 0.47 1.04 ± 0.24 0.91 ± 0.22 2.08 ± 2.92 1.24 ± 1.95 500 ± 669
b 0.0001a b 0.0001b 0.02b b 0.0001b b 0.0001b n.a. b 0.0001a 0.38a 0.0006a b 0.0001a b 0.0001a b 0.0001a b 0.0001a b 0.0001a b 0.0001a b 0.0001 b 0.0001 0.04
Fig. 1. ApoA-I (a), apoB (b) and apoM (c) concentration in plasma among patients and healthy control individuals (⁎⁎⁎p b 0.0001).
total cholesterol, HDL-c, apoA-I and apoB. Non-HDL-cholesterol (nonHDL-c) was defined as the difference between total cholesterol and HDL-c. Non-routine analyses, including interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α) and immunologic up-regulation analyzed with CD40 ligand (CD40L), were measured as previously described [20].
Table 2 Apolipoprotein levels in statin-treated and non-treated subjects given as mean ± SD. Protein ApoA-I (g/L)
ApoB (g/L)
ApoM (μmol/L)
Statin treated No statin treatment p-value Statin treated No statin treatment p-value Statin treated No statin treatment p-value
Using Mann–Whitney test.
Cases
Controls
p-value
1.60 ± 0.36 1.62 ± 0.34 0.3681 0.80 ± 0.19 0.99 ± 0.25 b 0.0001 0.67 ± 0.19 0.75 ± 0.22 0.0008
1.92 ± 0.39 2.10 ± 0.48 0.0378 0.86 ± 0.20 1.07 ± 0.23 b0.0001 0.85 ± 0.23 0.92 ± 0.21 0.1484
0.0001 b0.0001 0.068 0.004 b0.0001 b0.0001
J. Ahnström et al. / Clinical Biochemistry 43 (2010) 407–410
A sandwich ELISA for apoM based on two monoclonal antibodies, M42 and M58, was used to quantify apoM, as described earlier [15]. In the present study, the interassay coefficient of variation (CV) of the ELISA was 10.2% at the 100% level. Statistical analyses Data are summarized by means and standard deviations or percentages. Differences between two groups were tested by the Mann–Whitney U test. Calculating Spearman rank correlation coefficients separately for cases and controls assessed correlation of apoM with other risk factors. In order to further specify the univariate results, a multiple logistic regression analysis was used. All variables with p b 0.3 in the univariate analysis (Table 1) were entered in a multiple logistic regression model. One variable was then omitted at a time, starting with the variable with the highest p-value and stopped when all remaining variables had p b 0.1. The odds ratios (OR) are expressed per 1 standard deviation (SD) for continuous variables. The p-values b 0.05 were considered to indicate statistical significance. GraphPad Prism version 4, Instat 3 and R [21] were used for the statistical calculations. Results The study included 343 cases with AAA and 214 elderly apparently healthy control individuals drawn from the background population (Table 1). The distribution of apoA-I, apoB and apoM concentrations in both cases and the control individuals were essentially Gaussian (data not shown). The plasma levels of all three apolipoproteins were lower in cases (Table 1, Fig. 1). The difference was largest for apoA-I and apoM, both with approximately 20% lower levels in both male and female cases, as compared to the control individuals. ApoA-I and apoM were both higher in women than in men in cases (12 and 14%, respectively) as well as among the controls (17 and 12 %, respectively), whereas there was only a small gender difference in apoB levels found in cases (2%). None of the three measured apolipoproteins correlated with the size of the AAA (Table 3). A significantly larger proportion of cases were on statin treatment as compared to the background population. However, when cases and controls were separated into groups depending on statin treatment, the differences in apoA-I and apoM levels between cases and the controls remained statistically significant in both statin-treated and untreated subjects. In contrast, the difference in apoB levels between cases and control individuals was only observed among those that did
409
Table 4 Results from multivariate logistic regressions models.
Age⁎ (years) Current smoker Male gender Statin treated BMI (kg/m2)⁎ ApoA-I⁎ (g/L) ApoB⁎ (g/L) IL-6⁎ (pg/mL) TNF-α⁎ (pg/mL)
OR (95% CI)
p-value
SD
1.76 9.17 2.66 3.49 0.71 0.53 0.86 14.3 1.41
b 0.0001 b 0.0001 0.005 0.001 b 0.0001 b 0.0001 0.047 b 0.0001 0.007
6.50 n.a. n.a. n.a. 4.32 0.47 0.25 28.5 4.07
(1.48–2.10) (4.01–21.0) (1.34–5.28) (1.63–7.47) (0.61–0.82) (0.43–0.64) (0.75–0.998) (3.98–51.2) (1.10–1.81)
Other variables, which were included in the analysis but which were not significant, included diabetes status, diastolic BP, total cholesterol, HDL-c, non-HDL-c, apoM and CD40L. n.a., not applicable. ⁎ ORs expressed per 1 SD from continuous variables.
not receive statin treatment (Table 2). As expected, the statin treatment lowered the apoB levels in both cases and the controls with approximately 20%. ApoA-I was only affected by statin treatment in the group with controls, while apoM levels were affected amongst the cases (Table 2). However, these statistical calculations should be taken with caution, as the number of individuals in the control group taking statins was low. Plasma levels of apoM correlated positively to those of cholesterol, apoA-I and apoB (Table 3). There were no associations between apolipoprotein plasma levels and other known risk factors, apart from very weak negative associations to IL-6 for apoB and apoM and to TNF-α for apoM among the cases (Table 3) [20]. In a backward stepwise logistic regression analysis, apoA-I was associated with AAA, having an OR (95% CI) of 0.53 (0.43–0.64), whereas apoM was not associated with AAA (Table 4). Smoking status, a well-known risk factor, was as expected positively associated with AAA (Table 4). Surprisingly, BMI and apoB also showed significant OR below 1 (Table 4). ApoA-I and apoM correlated strongly and in a model without apoA-I, we found a significant OR for apoM of 0.78 (0.68–0.90) with p b 0.01, indicating that apoM is negatively associated to AAA but dependent on apoA-I. Discussion Even though many of the risk factors for AAA and atherosclerosis are the same, studies investigating the association between lipids/ lipoproteins and AAA have provided different results [2,4,5,8–10]. Out of seven independent population-based studies, total cholesterol was
Table 3 Results from correlation studies of apolipoproteins with different parameters (r-values). ApoA-I (g/L)
Age (years) AAA-max (mm) BMI (kg/m2) Systolic BP (mmHg) Diastolic BP (mmHg) Total cholesterol (mmol/L) HDL-c (mmol/L) Non-HDL-c (mmol/L) ApoA-I (g/L) ApoB (g/L) ApoM (μmol/L) CRP (mg/L) IL-6 (pg/mL) TNF-α (pg/mL) CD40L (pg/mL)
ApoB (g/L)
ApoM (μmol/L)
Cases (n = 343)
Controls (n = 214)
Cases (n = 343)
Controls (n = 214)
Cases (n = 343)
Controls (n = 214)
− 0.18⁎⁎ − 0.04 0.15⁎⁎ 0.01 0.06 0.23⁎⁎⁎ 0.62⁎⁎⁎ − 0.02 n.a. − 0.02 0.41⁎⁎⁎ − 0.04 − 0.12 − 0.06 − 0.07
0.03 n.a. − 0.16⁎ 0.04 0.03 0.17⁎ 0.56⁎⁎⁎ 0.01 n.a. − 0.13 0.40⁎⁎⁎ n.d. − 0.10 − 0.11 − 0.15
− 0.001 − 0.08 − 0.05 0.04 − 0.03 0.22⁎⁎⁎ 0.18⁎⁎ 0.81⁎⁎⁎ − 0.02 n.a. 0.39⁎⁎⁎ − 0.003 − 0.15⁎⁎ 0.002 0.02
0.34 n.a. − 0.01 0.02 − 0.04 0.55⁎⁎⁎ 0.15⁎ 0.52⁎⁎⁎ − 0.13 n.a. 0.24⁎⁎⁎ n.d. − 0.01 − 0.02 0.03
− 0.18⁎⁎⁎ − 0.08 − 0.10 0.14⁎⁎ 0.08 0.52⁎⁎⁎ 0.38⁎⁎⁎ 0.39⁎⁎⁎ 0.41⁎⁎⁎ 0.39⁎⁎⁎ n.a. − 0.03 −0.13⁎ − 0.17⁎⁎ − 0.09
− 0.04 n.a. − 0.15⁎ 0.002 0.11 0.31⁎⁎⁎ 0.49⁎⁎⁎ 0.19⁎⁎ 0.40⁎⁎⁎ 0.24⁎⁎⁎ n.a. n.d. − 0.16 − 0.12 − 0.12
Using Spearman rank-based correlation. n.a., not applicable. ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.0001.
410
J. Ahnström et al. / Clinical Biochemistry 43 (2010) 407–410
associated with AAA in only two [4,11]. LDL-c was identified as risk factor for AAA in three studies, whereas HDL-c was negatively associated with AAA in four studies [4,5,8,9,11]. Some of the discrepancies may result from methodological differences or from high proportions of statin-treated subjects in some study groups. In the only study measuring apoA-I and apoB, decreased apoA-I and increased apoB levels were reported [8]. ApoM has previously not been studied in AAA but has been suggested to have anti-atherosclerotic properties in mice [17,18] and in in vitro studies of isolated apoMcontaining human HDL [14,17]. Despite these results, in two prospective investigations, we have not been able to identify apoM as a predictor for CHD in humans [16]. In the present study, we investigated the plasma concentrations of apoA-I, apoB and apoM in AAA patients. To be able to identify risk factors of AAA, we decided to compare the AAA subjects to elderly healthy control individuals drawn from the background population. These individuals took part of a large population-based study investigating preventive factors for cardiovascular disease and alcohol abuse [19]. Many of the differences that we observed between the cases and the control individuals represent known risk factors of AAA, such as male gender and smoking suggesting that it was valid to use this population for comparisons. We demonstrate that apoA-I, apoB and apoM were all significantly lower in the cases than in the controls. We found higher apoA-I and apoM levels among women than in men, as has been described for apoA-I previously [22–24]. However, these differences were not large enough to explain the decrease of apoA-I and apoM in AAA. In multivariate analysis, only apoA-I and apoB were independently associated to AAA. It is not obvious why apoM levels are decreased in AAA and not in CHD [16], but it may reflect differences in pathologic disease progression. In this context, it is important to point out that the AAA patients had ongoing disease, whereas the two CHD studies were prospective and apoM was measured in samples drawn up to 20 years before the development of the disease [16]. The lower apoA-I levels in AAA patients were expected, but the low apoB levels were a surprising finding because apoB concentrations have previously been shown to be increased in AAA [8]. The low apoB levels were not possible to explain by the high proportion of statin treatment in the case group. ApoM levels were decreased in statin-treated AAA subjects. The reason for this decrease is not known but may be linked to the wellestablished strong correlation between apoM and total cholesterol levels. The mechanisms explaining this correlation are still unknown but as statin treatment directly affects the cholesterol levels, it is tempting to speculate that the decrease seen in apoM levels are due to decreased cholesterol levels. In agreement with previous observations, medium to strong positive correlations were found between levels of apoM and cholesterol, apoA-I and apoB, and this was not affected by the presence of disease [15,16]. The negative associations between apoM and IL-6 and TNF-α are too weak to allow conclusions to be drawn about the role of apoM in the inflammatory part of the disease progression. In multivariate analyses, apoA-I and apoB were associated with AAA in this population, whereas cholesterol levels were not. This suggests that investigations of apoA-I and apoB may be preferred to conventional lipid parameters in evaluation of AAA patients, just like in cardiovascular diseases [25]. The OR for apoM was not significant unless apoA-I was excluded from the model, indicating a strong association between apoM and apoA-I. Whether the lower apolipoprotein levels seen in the AAA group were caused by decreased synthesis or by increased elimination is unknown. In conclusion, in this population-based study of AAA, the plasma levels of apoA-I, apoB and apoM were significantly decreased in AAA cases as compared to the controls from the background population. ApoA-I and apoB were found to be independently associated with AAA. In contrast, the decreased OR associated with apoM was dependent on apoA-I.
Acknowledgments The authors would like to thank Ms. Helena Andersson for valuable technical assistance and Dr Michael Green for expert statistical assistance and discussions. This work was supported by grants from the Swedish Research Council (No. 07143), the Swedish Heart-Lung foundation, the Påhlsson's foundation, the Wallenberg Foundation, the Ernhold Lundström Foundation, the Hulda Almroth Foundation and research funds from the University Hospital in Malmö. References [1] Golledge J, Muller J, Daugherty A, Norman P. Abdominal aortic aneurysm: pathogenesis and implications for management. Arterioscler Thromb Vasc Biol 2006;26:2605–13. [2] Blanchard JF, Armenian HK, Friesen PP. Risk factors for abdominal aortic aneurysm: results of a case–control study. Am J Epidemiol 2000;151:575–83. [3] Lindblad B, Borner G, Gottsater A. Factors associated with development of large abdominal aortic aneurysm in middle-aged men. Eur J Vasc Endovasc Surg 2005;30:346–52. [4] Wanhainen A, Bergqvist D, Boman K, Nilsson TK, Rutegard J, Bjorck M. Risk factors associated with abdominal aortic aneurysm: a population-based study with historical and current data. J Vasc Surg 2005;41:390–6. [5] Hobbs SD, Claridge MW, Quick CR, Day NE, Bradbury AW, Wilmink AB. LDL cholesterol is associated with small abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2003;26:618–22. [6] Schlosser FJ, Tangelder MJ, Verhagen HJ, van der Heijden GJ, Muhs BE, van der Graaf Y, Moll FL. Growth predictors and prognosis of small abdominal aortic aneurysms. J Vasc Surg 2008;47:1127–33. [7] Choke E, Cockerill G, Wilson WR, Sayed S, Dawson J, Loftus I, Thompson MM. A review of biological factors implicated in abdominal aortic aneurysm rupture. Eur J Vasc Endovasc Surg 2005;30:227–44. [8] Simoni G, Gianotti A, Ardia A, Baiardi A, Galleano R, Civalleri D. Screening study of abdominal aortic aneurysm in a general population: lipid parameters. Cardiovasc Surg 1996;4:445–8. [9] Rizzo M, Krayenbuhl PA, Pernice V, Frasheri A, Battista Rini G, Berneis K. LDL size and subclasses in patients with abdominal aortic aneurysm. Int J Cardiol 2008. [10] Golledge J, Tsao PS, Dalman RL, Norman PE. Circulating markers of abdominal aortic aneurysm presence and progression. Circulation 2008;118:2382–92. [11] Forsdahl SH, Singh K, Solberg S, Jacobsen BK. Risk factors for abdominal aortic aneurysms: a 7-year prospective study: the Tromso Study, 1994–2001. Circulation 2009;119:2202–8. [12] Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med 1999;340:115–26. [13] Xu N, Dahlback B. A novel human apolipoprotein (apoM). J Biol Chem 1999;274: 31286–90. [14] Christoffersen C, Nielsen LB, Axler O, Andersson A, Johnsen AH, Dahlback B. Isolation and characterization of human apolipoprotein M-containing lipoproteins. J Lipid Res 2006;47:1833–43. [15] Axler O, Ahnstrom J, Dahlback B. An ELISA for apolipoprotein M reveals a strong correlation to total cholesterol in human plasma. J Lipid Res 2007;48:1772–80. [16] Ahnstrom J, et al. Levels of apolipoprotein M are not associated with the risk of coronary heart disease in two independent case–control studies. J Lipid Res 2008;49:1912–7. [17] Christoffersen C, Jauhiainen M, Moser M, Porse B, Ehnholm C, Boesl M, Dahlback B, Nielsen LB. Effect of apolipoprotein M on high density lipoprotein metabolism and atherosclerosis in low density lipoprotein receptor knock-out mice. J Biol Chem 2008;283:1839–47. [18] Wolfrum C, Poy MN, Stoffel M. Apolipoprotein M is required for prebeta-HDL formation and cholesterol efflux to HDL and protects against atherosclerosis. Nat Med 2005;11:418–22. [19] Nilsson P, Berglund G. Prevention of cardiovascular disease and diabetes: lessons from the Malmo Preventive Project. J Intern Med 2000;248:455–62. [20] Flondell-Site D, Lindblad B, Kolbel T, Gottsater A. Cytokines and systemic biomarkers are related to the size of abdominal aortic aneurysms. Cytokine 2009;46:211–5. [21] Team RDC. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Comuting; 2009. [22] Ritchie RF, Palomaki GE, Neveux LM, Ledue TB, Craig WY, Marcovina S, Navolotskaia O. Reference distributions for apolipoproteins AI and B and the apolipoprotein B/AI ratios: a practical and clinically relevant approach in a large cohort. J Clin Lab Anal 2006;20:209–17. [23] Contois J, McNamara JR, Lammi-Keefe C, Wilson PW, Massov T, Schaefer EJ. Reference intervals for plasma apolipoprotein A-1 determined with a standardized commercial immunoturbidimetric assay: results from the Framingham Offspring Study. Clin Chem 1996;42:507–14. [24] Jungner I, Marcovina SM, Walldius G, Holme I, Kolar W, Steiner E. Apolipoprotein B and A-I values in 147576 Swedish males and females, standardized according to the World Health Organization-International Federation of Clinical Chemistry First International Reference Materials. Clin Chem 1998;44:1641–9. [25] Andrikoula M, McDowell IF. The contribution of ApoB and ApoA1 measurements to cardiovascular risk assessment. Diabetes Obes Metab 2008;10:271–8.
Contents lists available at ScienceDirect
Clinical Biochemistry j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c l i n b i o c h e m
Plasma concentrations of apolipoproteins A-I, B and M in patients with abdominal aortic aneurysms Josefin Ahnström a, Anders Gottsäter b, Bengt Lindblad b, Björn Dahlbäck a,⁎ a b
Wallenberg Laboratory, Department of Laboratory Medicine, Clinical Chemistry, Lund University, SE-205 02 Malmö, Sweden Vascular Centre Malmö-Lund, Malmö University Hospital UMAS, Malmö, Sweden
a r t i c l e
i n f o
Article history: Received 6 August 2009 Received in revised form 22 October 2009 Accepted 10 November 2009 Available online 22 November 2009 Keywords: ApoM Signal peptide HDL Apolipoprotein Lipoprotein Lipocalin AAA ApoA-I ApoB Aneurysmal disease
a b s t r a c t Objectives: Apolipoproteins play important roles in the development of atherosclerosis but their involvement in the pathogenesis of abdominal aortic aneurysm (AAA) is poorly understood. The aim was to investigate whether apoA-I, apoB and apoM are independently associated with AAA. Design and methods: Plasma apoA-I, apoB and apoM were measured in 343 patients with AAA and in 214 elderly apparently healthy control individuals from the background population. Results: AAA patients had lower apolipoprotein levels, as compared to healthy individuals: apoA-I, 1.62 vs. 2.08 g/L; apoB, 0.91 vs. 1.04 g/L; apoM, 0.72 vs. 0.91 μmol/L (p b 0.0001 for all three). In multivariate analyses, apoA-I and apoB were associated with AAA, odds ratios (95% confidence intervals) being 0.53 (0.43–0.64) and 0.86 (0.75–0.998), respectively. Conclusions: ApoA-I, apoB and apoM levels were significantly lower in patients with AAA than in the control individuals, but only apoA-I and apoB were independently associated to AAA. © 2009 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Abdominal aortic aneurysm (AAA) has increasingly been recognized as an important cause of mortality with an annual mortality in the US of around 15,000 individuals. The prevalence in an aged population (N60 years of age) is around 2–4% [1–4]. AAA usually occurs in the infrarenal part of the aorta and is generally defined as a maximal aortic diameter of ≥3 cm [2,4–6]. The underlying problem in aneurysmal disease is weakening of the aortic wall, resulting in progressive dilation and, if left untreated, risk for eventual aneurysm rupture. The principal determinant of AAA rupture is the maximum diameter of the aneurysm [7]. Male gender, smoking, hypertension, increasing age, coronary heart disease (CHD) and a family history of abdominal aortic aneurysm have been identified as independent risk factors for AAA in several population-based studies [1–5]. However, data on the association between lipid abnormalities and the risk of AAA are conflicting [2,4,5,8–11], even though it is well recognized that oxidized lowdensity lipoprotein (LDL) is a major cause of injury to vascular
⁎ Corresponding author. Wallenberg Laboratory, Division of Clinical Chemistry, Department of Laboratory Medicine, University Hospital, Lund University, Entrance 46, Floor 6, Malmö S-20502 Malmö, Sweden. Fax: +46 40 337044. E-mail address: [email protected] (B. Dahlbäck).
endothelium and smooth muscle cells in the media [12]. In CHD, apoA-I and apoB, the major apolipoproteins of high density lipoproteins (HDL) and LDL, respectively, have been shown to be equivalent or better than the HDL-cholesterol (HDL-c) and LDLcholesterol (LDL-c) as risk markers for atherosclerosis or cardiovascular events. ApoA-I and apoB have, as far as we know, only been measured in one study of AAA, which demonstrated decreased apoA-I and increased apoB levels in the patients [8]. The relatively recently discovered apolipoprotein M (apoM) is preferentially bound to HDL but it is also, to a minor extent, associated with the other lipoproteins [13,14]. ApoM is present in around 5% of HDL and in 2% of LDL particles [14]. The apoM concentration in human plasma is approximately 0.9 μmol/L (∼23 mg/L) with somewhat lower levels in young women than in older women and men [15]. It correlates positively to total plasma cholesterol and to the plasma concentrations of both LDL- and HDL-cholesterol [15,16]. In vivo studies in mice suggested apoM to play a role in HDL metabolism and to reduce the development of atherosclerosis [17,18]. Studies on isolated apoM-containing human HDL particles showed that those particles were more efficient in reducing LDL oxidation and stimulating cholesterol efflux than apoM-free HDL [14]. Despite the strong correlation of plasma apoM to total plasma cholesterol concentrations, and the positive results seen in mice, two independent case–control studies of CHD did not identify apoM as a predictor of disease [16].
0009-9120/$ – see front matter © 2009 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2009.11.006
408
J. Ahnström et al. / Clinical Biochemistry 43 (2010) 407–410
Aortic aneurysm disease shares many risk factors with cardiovascular diseases, but it is a distinct entity with unique pathogenic mechanisms. The aim of this study was to investigate whether there is a relationship between plasma apoA-I, apoB and apoM levels and the development of AAA. Materials and methods Study design AAA patients undergoing routine ultrasound follow-up at the Vascular Centre, Malmö University Hospital from February 2002 to December 2006 were included in the study. All investigations and blood sampling were performed prior to any invasive treatment. No patients with acute symptomatic or ruptured AAA were included. The study included 343 patients with either large AAA planned for operation or small AAA undergoing surveillance. The AAA patients were compared to individuals drawn from a background population selected from a large population study investigating preventive factors for cardiovascular disease and alcohol abuse [19]. The individuals, consisting of 214 elderly apparently healthy individuals without symptomatic AAA, cardiovascular disease or atherosclerosis, were included during 2004–2005 and were chosen to represent the general population of approximately the same age as the patients. Further data are given in Table 1. The Ethical Committee of Lund University approved the study and all patients gave written consent to participate. Measurements At baseline, all patients were evaluated regarding blood pressure (BP) (measured in the right arm after 10 min rest), diabetes mellitus (treatment of diabetes or a fasting blood glucose level of ≥6.7 mmol/L) and pharmacological treatment with cardiovascular drugs. Body mass index (BMI) was calculated as kg/m2. Smoking was defined as current smoking. Blood sampling was performed at baseline in all participants. Blood samples in appropriate tubes were directly handled and centrifuged (at 4 °C). Routine laboratory analyses were performed immediately. Samples for non-routine analyses were frozen in −80 °C for later analysis. Standard laboratory methods at the Department of Clinical Chemistry, Malmö University Hospital (SWEDAC approved according to European norm 45001), were used for analyses including
Table 1 Data for the subjects included in the study given as mean ± SD or proportion.
Age (years) Men (%) Diabetes mellitus (%) Smokers (%) Statin treatment (%) AAA-max (mm) BMI (kg/m2) Systolic BP (mmHg) Diastolic BP (mmHg) Cholesterol (mmol/L) HDL-c (mmol/L) Non-HDL-c (mmol/L) ApoA-I (g/L) ApoB (g/L) ApoM (μmol/L) IL-6 (pg/mL) TNF-α (pg/mL) CD40L (pg/mL) n.a., not applicable. a Mann–Whitney test. b χ2 test.
Cases (n = 343)
Controls (n = 214)
p-value
74 ± 8 79 11.9 34.6 38.7 62.8 ± 14.6 25.4 ± 4.05 141.1 ± 19.4 81.3 ± 10.5 4.98 ± 1.22 1.06 ± 0.35 3.95 ± 1.21 1.62 ± 0.36 0.91 ± 0.25 0.72 ± 0.21 9.36 ± 32.19 2.73 ± 6.56 582 ± 2229
68 ± 2 46.3 5.7 12.2 13.1 n.a. 27.1 ± 4.33 143.4 ± 19.7 84.9 ± 10.4 5.85 ± 1.07 1.24 ± 0.34 4.59 ± 1.02 2.08 ± 0.47 1.04 ± 0.24 0.91 ± 0.22 2.08 ± 2.92 1.24 ± 1.95 500 ± 669
b 0.0001a b 0.0001b 0.02b b 0.0001b b 0.0001b n.a. b 0.0001a 0.38a 0.0006a b 0.0001a b 0.0001a b 0.0001a b 0.0001a b 0.0001a b 0.0001a b 0.0001 b 0.0001 0.04
Fig. 1. ApoA-I (a), apoB (b) and apoM (c) concentration in plasma among patients and healthy control individuals (⁎⁎⁎p b 0.0001).
total cholesterol, HDL-c, apoA-I and apoB. Non-HDL-cholesterol (nonHDL-c) was defined as the difference between total cholesterol and HDL-c. Non-routine analyses, including interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α) and immunologic up-regulation analyzed with CD40 ligand (CD40L), were measured as previously described [20].
Table 2 Apolipoprotein levels in statin-treated and non-treated subjects given as mean ± SD. Protein ApoA-I (g/L)
ApoB (g/L)
ApoM (μmol/L)
Statin treated No statin treatment p-value Statin treated No statin treatment p-value Statin treated No statin treatment p-value
Using Mann–Whitney test.
Cases
Controls
p-value
1.60 ± 0.36 1.62 ± 0.34 0.3681 0.80 ± 0.19 0.99 ± 0.25 b 0.0001 0.67 ± 0.19 0.75 ± 0.22 0.0008
1.92 ± 0.39 2.10 ± 0.48 0.0378 0.86 ± 0.20 1.07 ± 0.23 b0.0001 0.85 ± 0.23 0.92 ± 0.21 0.1484
0.0001 b0.0001 0.068 0.004 b0.0001 b0.0001
J. Ahnström et al. / Clinical Biochemistry 43 (2010) 407–410
A sandwich ELISA for apoM based on two monoclonal antibodies, M42 and M58, was used to quantify apoM, as described earlier [15]. In the present study, the interassay coefficient of variation (CV) of the ELISA was 10.2% at the 100% level. Statistical analyses Data are summarized by means and standard deviations or percentages. Differences between two groups were tested by the Mann–Whitney U test. Calculating Spearman rank correlation coefficients separately for cases and controls assessed correlation of apoM with other risk factors. In order to further specify the univariate results, a multiple logistic regression analysis was used. All variables with p b 0.3 in the univariate analysis (Table 1) were entered in a multiple logistic regression model. One variable was then omitted at a time, starting with the variable with the highest p-value and stopped when all remaining variables had p b 0.1. The odds ratios (OR) are expressed per 1 standard deviation (SD) for continuous variables. The p-values b 0.05 were considered to indicate statistical significance. GraphPad Prism version 4, Instat 3 and R [21] were used for the statistical calculations. Results The study included 343 cases with AAA and 214 elderly apparently healthy control individuals drawn from the background population (Table 1). The distribution of apoA-I, apoB and apoM concentrations in both cases and the control individuals were essentially Gaussian (data not shown). The plasma levels of all three apolipoproteins were lower in cases (Table 1, Fig. 1). The difference was largest for apoA-I and apoM, both with approximately 20% lower levels in both male and female cases, as compared to the control individuals. ApoA-I and apoM were both higher in women than in men in cases (12 and 14%, respectively) as well as among the controls (17 and 12 %, respectively), whereas there was only a small gender difference in apoB levels found in cases (2%). None of the three measured apolipoproteins correlated with the size of the AAA (Table 3). A significantly larger proportion of cases were on statin treatment as compared to the background population. However, when cases and controls were separated into groups depending on statin treatment, the differences in apoA-I and apoM levels between cases and the controls remained statistically significant in both statin-treated and untreated subjects. In contrast, the difference in apoB levels between cases and control individuals was only observed among those that did
409
Table 4 Results from multivariate logistic regressions models.
Age⁎ (years) Current smoker Male gender Statin treated BMI (kg/m2)⁎ ApoA-I⁎ (g/L) ApoB⁎ (g/L) IL-6⁎ (pg/mL) TNF-α⁎ (pg/mL)
OR (95% CI)
p-value
SD
1.76 9.17 2.66 3.49 0.71 0.53 0.86 14.3 1.41
b 0.0001 b 0.0001 0.005 0.001 b 0.0001 b 0.0001 0.047 b 0.0001 0.007
6.50 n.a. n.a. n.a. 4.32 0.47 0.25 28.5 4.07
(1.48–2.10) (4.01–21.0) (1.34–5.28) (1.63–7.47) (0.61–0.82) (0.43–0.64) (0.75–0.998) (3.98–51.2) (1.10–1.81)
Other variables, which were included in the analysis but which were not significant, included diabetes status, diastolic BP, total cholesterol, HDL-c, non-HDL-c, apoM and CD40L. n.a., not applicable. ⁎ ORs expressed per 1 SD from continuous variables.
not receive statin treatment (Table 2). As expected, the statin treatment lowered the apoB levels in both cases and the controls with approximately 20%. ApoA-I was only affected by statin treatment in the group with controls, while apoM levels were affected amongst the cases (Table 2). However, these statistical calculations should be taken with caution, as the number of individuals in the control group taking statins was low. Plasma levels of apoM correlated positively to those of cholesterol, apoA-I and apoB (Table 3). There were no associations between apolipoprotein plasma levels and other known risk factors, apart from very weak negative associations to IL-6 for apoB and apoM and to TNF-α for apoM among the cases (Table 3) [20]. In a backward stepwise logistic regression analysis, apoA-I was associated with AAA, having an OR (95% CI) of 0.53 (0.43–0.64), whereas apoM was not associated with AAA (Table 4). Smoking status, a well-known risk factor, was as expected positively associated with AAA (Table 4). Surprisingly, BMI and apoB also showed significant OR below 1 (Table 4). ApoA-I and apoM correlated strongly and in a model without apoA-I, we found a significant OR for apoM of 0.78 (0.68–0.90) with p b 0.01, indicating that apoM is negatively associated to AAA but dependent on apoA-I. Discussion Even though many of the risk factors for AAA and atherosclerosis are the same, studies investigating the association between lipids/ lipoproteins and AAA have provided different results [2,4,5,8–10]. Out of seven independent population-based studies, total cholesterol was
Table 3 Results from correlation studies of apolipoproteins with different parameters (r-values). ApoA-I (g/L)
Age (years) AAA-max (mm) BMI (kg/m2) Systolic BP (mmHg) Diastolic BP (mmHg) Total cholesterol (mmol/L) HDL-c (mmol/L) Non-HDL-c (mmol/L) ApoA-I (g/L) ApoB (g/L) ApoM (μmol/L) CRP (mg/L) IL-6 (pg/mL) TNF-α (pg/mL) CD40L (pg/mL)
ApoB (g/L)
ApoM (μmol/L)
Cases (n = 343)
Controls (n = 214)
Cases (n = 343)
Controls (n = 214)
Cases (n = 343)
Controls (n = 214)
− 0.18⁎⁎ − 0.04 0.15⁎⁎ 0.01 0.06 0.23⁎⁎⁎ 0.62⁎⁎⁎ − 0.02 n.a. − 0.02 0.41⁎⁎⁎ − 0.04 − 0.12 − 0.06 − 0.07
0.03 n.a. − 0.16⁎ 0.04 0.03 0.17⁎ 0.56⁎⁎⁎ 0.01 n.a. − 0.13 0.40⁎⁎⁎ n.d. − 0.10 − 0.11 − 0.15
− 0.001 − 0.08 − 0.05 0.04 − 0.03 0.22⁎⁎⁎ 0.18⁎⁎ 0.81⁎⁎⁎ − 0.02 n.a. 0.39⁎⁎⁎ − 0.003 − 0.15⁎⁎ 0.002 0.02
0.34 n.a. − 0.01 0.02 − 0.04 0.55⁎⁎⁎ 0.15⁎ 0.52⁎⁎⁎ − 0.13 n.a. 0.24⁎⁎⁎ n.d. − 0.01 − 0.02 0.03
− 0.18⁎⁎⁎ − 0.08 − 0.10 0.14⁎⁎ 0.08 0.52⁎⁎⁎ 0.38⁎⁎⁎ 0.39⁎⁎⁎ 0.41⁎⁎⁎ 0.39⁎⁎⁎ n.a. − 0.03 −0.13⁎ − 0.17⁎⁎ − 0.09
− 0.04 n.a. − 0.15⁎ 0.002 0.11 0.31⁎⁎⁎ 0.49⁎⁎⁎ 0.19⁎⁎ 0.40⁎⁎⁎ 0.24⁎⁎⁎ n.a. n.d. − 0.16 − 0.12 − 0.12
Using Spearman rank-based correlation. n.a., not applicable. ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.0001.
410
J. Ahnström et al. / Clinical Biochemistry 43 (2010) 407–410
associated with AAA in only two [4,11]. LDL-c was identified as risk factor for AAA in three studies, whereas HDL-c was negatively associated with AAA in four studies [4,5,8,9,11]. Some of the discrepancies may result from methodological differences or from high proportions of statin-treated subjects in some study groups. In the only study measuring apoA-I and apoB, decreased apoA-I and increased apoB levels were reported [8]. ApoM has previously not been studied in AAA but has been suggested to have anti-atherosclerotic properties in mice [17,18] and in in vitro studies of isolated apoMcontaining human HDL [14,17]. Despite these results, in two prospective investigations, we have not been able to identify apoM as a predictor for CHD in humans [16]. In the present study, we investigated the plasma concentrations of apoA-I, apoB and apoM in AAA patients. To be able to identify risk factors of AAA, we decided to compare the AAA subjects to elderly healthy control individuals drawn from the background population. These individuals took part of a large population-based study investigating preventive factors for cardiovascular disease and alcohol abuse [19]. Many of the differences that we observed between the cases and the control individuals represent known risk factors of AAA, such as male gender and smoking suggesting that it was valid to use this population for comparisons. We demonstrate that apoA-I, apoB and apoM were all significantly lower in the cases than in the controls. We found higher apoA-I and apoM levels among women than in men, as has been described for apoA-I previously [22–24]. However, these differences were not large enough to explain the decrease of apoA-I and apoM in AAA. In multivariate analysis, only apoA-I and apoB were independently associated to AAA. It is not obvious why apoM levels are decreased in AAA and not in CHD [16], but it may reflect differences in pathologic disease progression. In this context, it is important to point out that the AAA patients had ongoing disease, whereas the two CHD studies were prospective and apoM was measured in samples drawn up to 20 years before the development of the disease [16]. The lower apoA-I levels in AAA patients were expected, but the low apoB levels were a surprising finding because apoB concentrations have previously been shown to be increased in AAA [8]. The low apoB levels were not possible to explain by the high proportion of statin treatment in the case group. ApoM levels were decreased in statin-treated AAA subjects. The reason for this decrease is not known but may be linked to the wellestablished strong correlation between apoM and total cholesterol levels. The mechanisms explaining this correlation are still unknown but as statin treatment directly affects the cholesterol levels, it is tempting to speculate that the decrease seen in apoM levels are due to decreased cholesterol levels. In agreement with previous observations, medium to strong positive correlations were found between levels of apoM and cholesterol, apoA-I and apoB, and this was not affected by the presence of disease [15,16]. The negative associations between apoM and IL-6 and TNF-α are too weak to allow conclusions to be drawn about the role of apoM in the inflammatory part of the disease progression. In multivariate analyses, apoA-I and apoB were associated with AAA in this population, whereas cholesterol levels were not. This suggests that investigations of apoA-I and apoB may be preferred to conventional lipid parameters in evaluation of AAA patients, just like in cardiovascular diseases [25]. The OR for apoM was not significant unless apoA-I was excluded from the model, indicating a strong association between apoM and apoA-I. Whether the lower apolipoprotein levels seen in the AAA group were caused by decreased synthesis or by increased elimination is unknown. In conclusion, in this population-based study of AAA, the plasma levels of apoA-I, apoB and apoM were significantly decreased in AAA cases as compared to the controls from the background population. ApoA-I and apoB were found to be independently associated with AAA. In contrast, the decreased OR associated with apoM was dependent on apoA-I.
Acknowledgments The authors would like to thank Ms. Helena Andersson for valuable technical assistance and Dr Michael Green for expert statistical assistance and discussions. This work was supported by grants from the Swedish Research Council (No. 07143), the Swedish Heart-Lung foundation, the Påhlsson's foundation, the Wallenberg Foundation, the Ernhold Lundström Foundation, the Hulda Almroth Foundation and research funds from the University Hospital in Malmö. References [1] Golledge J, Muller J, Daugherty A, Norman P. Abdominal aortic aneurysm: pathogenesis and implications for management. Arterioscler Thromb Vasc Biol 2006;26:2605–13. [2] Blanchard JF, Armenian HK, Friesen PP. Risk factors for abdominal aortic aneurysm: results of a case–control study. Am J Epidemiol 2000;151:575–83. [3] Lindblad B, Borner G, Gottsater A. Factors associated with development of large abdominal aortic aneurysm in middle-aged men. Eur J Vasc Endovasc Surg 2005;30:346–52. [4] Wanhainen A, Bergqvist D, Boman K, Nilsson TK, Rutegard J, Bjorck M. Risk factors associated with abdominal aortic aneurysm: a population-based study with historical and current data. J Vasc Surg 2005;41:390–6. [5] Hobbs SD, Claridge MW, Quick CR, Day NE, Bradbury AW, Wilmink AB. LDL cholesterol is associated with small abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2003;26:618–22. [6] Schlosser FJ, Tangelder MJ, Verhagen HJ, van der Heijden GJ, Muhs BE, van der Graaf Y, Moll FL. Growth predictors and prognosis of small abdominal aortic aneurysms. J Vasc Surg 2008;47:1127–33. [7] Choke E, Cockerill G, Wilson WR, Sayed S, Dawson J, Loftus I, Thompson MM. A review of biological factors implicated in abdominal aortic aneurysm rupture. Eur J Vasc Endovasc Surg 2005;30:227–44. [8] Simoni G, Gianotti A, Ardia A, Baiardi A, Galleano R, Civalleri D. Screening study of abdominal aortic aneurysm in a general population: lipid parameters. Cardiovasc Surg 1996;4:445–8. [9] Rizzo M, Krayenbuhl PA, Pernice V, Frasheri A, Battista Rini G, Berneis K. LDL size and subclasses in patients with abdominal aortic aneurysm. Int J Cardiol 2008. [10] Golledge J, Tsao PS, Dalman RL, Norman PE. Circulating markers of abdominal aortic aneurysm presence and progression. Circulation 2008;118:2382–92. [11] Forsdahl SH, Singh K, Solberg S, Jacobsen BK. Risk factors for abdominal aortic aneurysms: a 7-year prospective study: the Tromso Study, 1994–2001. Circulation 2009;119:2202–8. [12] Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med 1999;340:115–26. [13] Xu N, Dahlback B. A novel human apolipoprotein (apoM). J Biol Chem 1999;274: 31286–90. [14] Christoffersen C, Nielsen LB, Axler O, Andersson A, Johnsen AH, Dahlback B. Isolation and characterization of human apolipoprotein M-containing lipoproteins. J Lipid Res 2006;47:1833–43. [15] Axler O, Ahnstrom J, Dahlback B. An ELISA for apolipoprotein M reveals a strong correlation to total cholesterol in human plasma. J Lipid Res 2007;48:1772–80. [16] Ahnstrom J, et al. Levels of apolipoprotein M are not associated with the risk of coronary heart disease in two independent case–control studies. J Lipid Res 2008;49:1912–7. [17] Christoffersen C, Jauhiainen M, Moser M, Porse B, Ehnholm C, Boesl M, Dahlback B, Nielsen LB. Effect of apolipoprotein M on high density lipoprotein metabolism and atherosclerosis in low density lipoprotein receptor knock-out mice. J Biol Chem 2008;283:1839–47. [18] Wolfrum C, Poy MN, Stoffel M. Apolipoprotein M is required for prebeta-HDL formation and cholesterol efflux to HDL and protects against atherosclerosis. Nat Med 2005;11:418–22. [19] Nilsson P, Berglund G. Prevention of cardiovascular disease and diabetes: lessons from the Malmo Preventive Project. J Intern Med 2000;248:455–62. [20] Flondell-Site D, Lindblad B, Kolbel T, Gottsater A. Cytokines and systemic biomarkers are related to the size of abdominal aortic aneurysms. Cytokine 2009;46:211–5. [21] Team RDC. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Comuting; 2009. [22] Ritchie RF, Palomaki GE, Neveux LM, Ledue TB, Craig WY, Marcovina S, Navolotskaia O. Reference distributions for apolipoproteins AI and B and the apolipoprotein B/AI ratios: a practical and clinically relevant approach in a large cohort. J Clin Lab Anal 2006;20:209–17. [23] Contois J, McNamara JR, Lammi-Keefe C, Wilson PW, Massov T, Schaefer EJ. Reference intervals for plasma apolipoprotein A-1 determined with a standardized commercial immunoturbidimetric assay: results from the Framingham Offspring Study. Clin Chem 1996;42:507–14. [24] Jungner I, Marcovina SM, Walldius G, Holme I, Kolar W, Steiner E. Apolipoprotein B and A-I values in 147576 Swedish males and females, standardized according to the World Health Organization-International Federation of Clinical Chemistry First International Reference Materials. Clin Chem 1998;44:1641–9. [25] Andrikoula M, McDowell IF. The contribution of ApoB and ApoA1 measurements to cardiovascular risk assessment. Diabetes Obes Metab 2008;10:271–8.