Complement C3 and high risk of venous thromboembolism: 80517 individuals from the Copenhagen General Population Study.
Complement C3 is an acute-phase reactant and a central component in activation of the complement system. The complement system is activated during blood clotting, suggesting participation of these proteins in thrombus formation (12). Several studies have suggested that complement activation appears to be involved in the pathophysiology of a diverse range of cardiac conditions (13, 14), and 1 study proposed that complement activation is associated with venous thrombosis in patients with systemic lupus erythematosus (15), a condition characterized by formation of immune complexes and inflammation in several organ systems. However, it is unknown whether high complement C3 concentrations are associated with high risk of venous thromboembolism in individuals in the general population.
We tested the hypothesis that high complement C3 concentrations are associated with high risk of venous thromboembolism in the general population. For this purpose, we measured baseline plasma concentrations of complement C3 in 80 517 white Danish individuals without venous thromboembolism from the Copenhagen General Population Study and followed them for up to 10 years, during which time 1176 developed venous thromboembolism, 739 deep venous thrombosis, and 529 pulmonary embolism.
The study was approved by Herlev and Gentofte Hospital, Copenhagen University Hospital, and Danish ethics committees and was conducted according to the Declaration of Helsinki. All participants gave written informed consent.
The Copenhagen General Population Study is aprospective study of the Danish general population initiated in 2003 (8). Participants were selected randomly from the national Danish Civil Registration System, which uniquely identifies all inhabitants of Denmark, to represent the general population: participants were whites of Danish descent, defined according to the Danish Civil Registration System, and the participant and both parents were required to be Danish citizens born in Denmark. All participants, aged 40-100 years, were from the greater Copenhagen area.
We obtained data from blood samples, physical examination, and self-administered questionnaire reviewed with an investigator on the day of attendance. Of 82 431 individuals examined from November 2003 through July 2012, we excluded 1914 with a diagnosis of venous thromboembolism before the examination date, leaving 80 517 individuals for analyses.
VENOUS THROMBOEMBOLISM (DEEP VENOUS THROMBOSIS AND PULMONARY EMBOLISM)
We followed all individuals from the day of examination until the occurrence of deep venous thrombosis, pulmonary embolism, death (n = 3880), emigration (n = 205), or end of follow-up in April 2013, whichever came first. No individuals were lost to follow-up. Diagnoses of deep venous thrombosis and pulmonary embolism were identified through the national Danish Patients Registry that contains information from all clinical hospital departments in Denmark from 1977 through April 2013, including outpatients and emergency wards from 1995, on the basis of International Classification of Diseases, Revision 8 and Revision 10 (ICD-8 and -10)  [deep venous thrombosis: 451.00, 451.08-09, 451.90, 451.92, 671.01-03, and 671.08-09 (ICD-8) and I80.1-3, O22.3, and O87.1 (ICD-10); pulmonary embolism: 450.99 and 673.99 (ICD-8) and I26.0, I26.9, and O88.2 (ICD-10)]. These diagnoses with identical ICD-8 and ICD-10 codes were also identified in the national Danish Causes of Death registry from 1977 through April 2013. This approach has been used previously (8, 9, 16) and was validated in 2010 with a positive predictive value of 59% for venous thromboembolism diagnosed in any setting and 75% for venous thromboembolism diagnosed during hospital admission (17). Information on death and emigration were obtained from the Danish Civil Registration System.
We measured plasma complement C3 concentrations on fresh samples with turbidimetry (Konelab) immediately after sampling. Measurement of complement C3 was blinded to later development of venous thromboembolism and vice versa. Internal controls were analyzed daily over the 9-year period, with typical monthly CVs of 3%-5% and no significant changes in the values.
Smoking status was divided into never, current, or former; daily tobacco smoking was measured by number of cigarettes or equivalent smoked at time of examination; and pack-years of cumulative smoking was calculated as 20 cigarettes or equivalents smoked daily for 1 year. Alcohol use was based on number of drinks per week (1 drink [approximately equals] 12 g alcohol). We also recorded the use of oral contraceptives among premenopausal women, hormone replacement therapy among postmenopausal women, marital status (single, married, separated, divorced, or widowed), level of education (elementary, 1-9 years of education; high school, 10-12 years; or academic, >12 years), and leisure-time physical activity (<2 h exercise per week, 2-4 h light exercise per week, 2-4 h demanding exercise per week, or >4 h exercise per week).
Height, weight, and systolic blood pressure were measured; body mass index (BMI) was calculated as weight divided by height squared. Plasma concentrations of total cholesterol, HDL cholesterol, LDL cholesterol, high-sensitivity C-reactive protein (CRP), rheumatoid factor, coagulation factors II + VII + X, international normalized ratio, fibrinogen, and glucose were measured at baseline with standard hospital assays. We genotyped for factor V Leiden R506Q (rs6025) and prothrombin G20210A (rs1799963) with TaqMan assays. Information on hospital admissions, major surgery, and cancer diagnoses were from the national Danish Patient Registry and the national Danish Cancer Registry.
We used Stata 13.1 statistical software. Information on covariates used for adjustments was 97.8% complete. Missing values for adjustment were imputed by multivariable normal regression imputation with age, sex, date of birth, and education level as predictors; however, if we included only individuals with complete data, the results were similar to those reported. In sensitivity analyses, we further adjusted for rheumatoid factor, which was measured on 43 944 individuals, and for factor V Leiden R506Q (rs6025) and prothrombin G20210A (rs1799963), which were genotyped in 51 025 individuals. We used the nptrend function in Stata to test for trend across tertiles of complement C3.
Unadjusted and CRP-adjusted complement C3 concentrations as a function of categories of BMI (<18.5; 18.5-24.9; 25-29.9; 30-34.9; 35-39.9; [greater than or equal to]40 kg/m2) and unadjusted and BMI-adjusted complement C3 concentrations as a function of categories of CRP concentrations (<1; 1-1.9; 2-2.9; 3-4.9; 5-9.9; [greater than or equal to]10 mg/L) were compared with the Kruskal-Wallis test; adjustment for CRP was performed because acute phase may influence complement C3 concentrations (18), whereas adjustment for BMI was done because adipose tissue produces complement C3 (19).
Cumulative incidences were calculated with competing risk regression, by use of a nonparametric method summing up to t from S(t - 1) X h'(t), where S(t - 1) is the Kaplan-Meier estimator of the overall survival function and h'(t) is the cause-specific hazard at time t, taking competing events into account. Standard errors were computed according to Marubini and Valsecchi (20). Log-rank trend tests examined differences between individuals in different complement C3 tertiles. In all analyses, competing events were death from other causes: in analyses of venous thromboembolism, 3601 competing deaths occurred; in analyses of deep venous thrombosis, 3766 competing deaths occurred; and in analyses of pulmonary embolism, 3673 competing deaths occurred.
Hazard ratios (HRs) with 95% CIs were calculated by use ofCox regression models, with age as time scale (referred to as age-adjusted) and delayed entry at examination. Multivariable-adjusted Cox models included known risk factors for venous thromboembolism, other lifestyle factors, and markers of social status at study entry; that is, age, sex, current smoking, pack-years of cumulative smoking, daily tobacco smoking, weekly alcohol consumption, use of oral contraceptives in premenopausal women, use of hormone replacement therapy in postmenopausal women, leisuretime physical activity, marital status, length of education, systolic blood pressure, total cholesterol, LDL cholesterol, HDL cholesterol, coagulation factors II + VII + X, international normalized ratio, fibrinogen, glucose, any cancer 1 year before or after an event, hospitalization within 3 months before an event, and major surgery within 3 months before an event. Because CRP is another acute-phase reactant associated with venous thromboembolism (11), we also calculated HRs additionally adjusted for CRP.
Finally, we calculated HRs additionally adjusted for BMI, because adipose tissue is a producer of complement C3 (19), lowering the amount of adipose tissue is associated with lower concentrations of C3 (21), and because overweight is a well-established causal risk factor for venous thromboembolism (22, 23).
Stratum-specific HRs of venous thromboembolism were calculated by performing subgroup analyses according to sex, age, BMI, smoking status, and CRP concentrations, which are important, common predictors ofvenous thromboembolism. For this purpose, we used plasma concentrations of complement C3 on a continuous scale. Tests for interaction were performed by introducing a 2-factor interaction term in Cox models.
During393 144person-years offollowup, 1176individuals developed venous thromboembolism, 739 deep venous thrombosis, and 529 pulmonary embolism. Plasma complement C3 concentrations were associated with many baseline characteristics of individuals, including CRP and BMI (Table 1). Complement C3 concentrations were approximately normally distributed (Supplemental Fig. 1, which accompanies the online version of this article at http://www.clinchem.org/content/vol62/ issue3), with a mean value of 1.13 g/L (interquartile range 0.98-1.26; SD 0.21).
COMPLEMENT C3, CRP, AND BMI
Complement C3 concentrations were higher as a function of progressively higher categories of BMI (Fig. 1, top left). This association was attenuated only slightly after adjustment for CRP concentrations (Fig. 1, bottom left). Complement C3 concentrations were also higher as a function of higher categories of CRP concentrations (Fig. 1, top right). This association was attenuated somewhat after adjustment for BMI (Fig. 1D, bottom right). Associations and correlations among complement C3 and primary predictors of venous thromboembolism are given in Table 2.
CUMULATIVE INCIDENCE BY COMPLEMENT C3 TERTILES
The cumulative incidences ofvenous thromboembolism, deep venous thrombosis, and pulmonary embolism as a function of age were higher with progressively higher tertiles of complement C3 concentrations (log-rank trend P = 3 X [10.sup.-8], 1 X [10.sup.-4], and 2 X [10.sup.-6], respectively) (Fig. 2): at age 80, 7%, 9%, and 11% of individuals in the first, second, and third tertiles, respectively, had developed venous thromboembolism.
VENOUS THROMBOEMBOLISM BY COMPLEMENT C3 TERTILES
Multivariable-adjusted HRs for venous thromboembolism compared with individuals in the first tertile were 1.36 (95% CI, 1.16-1.59) for those in the second tertile and 1.58 (1.33-1.88) for those in the third tertile (Fig. 3, upper panel). Corresponding values were 1.36 (1.16-1.60) and 1.57 (1.33-1.87) after additional adjustment for CRP and 1.27 (1.09-1.49) and 1.31 (1.10-1.57) after additional adjustment for BMI.
DEEP VENOUS THROMBOSIS BY COMPLEMENT C3 TERTILES
Multivariable-adjusted HRs for deep venous thrombosis compared with individuals in the first tertile were 1.41 (95% CI, 1.16-1.72) for those in the second tertile and 1.66 (1.33-2.05) for those in the third tertile (Fig. 3, middle panel). Corresponding values were 1.41 (1.16-1.72) and 1.65 (1.33-2.04) after additional adjustment for CRP and 1.29 (1.06-1.57) and 1.28 (1.02-1.60) after additional adjustment for BMI.
PULMONARY EMBOLISM BY COMPLEMENT C3 TERTILES
Multivariable-adjusted HRs for pulmonary embolism compared with individuals in the first tertile were 1.35 (95% CI, 1.06-1.71) for those in the second tertile and 1.66 (1.28-2.14) for those in the third tertile (Fig. 3, lower panel). Corresponding values were 1.35 (1.06-1.71) and 1.65 (1.28-2.14) after additional adjustment for CRP and 1.30 (1.02-1.65) and 1.49 (1.14-1.95) after additional adjustment for BMI.
When further adjusting for rheumatoid factor, which was measured on 43 944 individuals or for factor V Leiden R506Q (rs6025) and prothrombin G20210A (rs1799963), which were genotyped in 51 025 individuals, HRs for venous thromboembolism, deep venous thrombosis, and pulmonary embolism were similar (compare online Supplemental Figs. 2 and 3 with Fig. 3). After removing the adjustment for cancer 1 year before and after an event from the model, results for complement C3 and risk of venous thromboembolism, deep venous thromboembolism, and pulmonary embolism were similar (compare online Supplemental Fig. 4 with Fig. 3).
VENOUS THROMBOSIS BY COMPLEMENT C3 ON A CONTINUOUS SCALE
The multivariable-adjusted HR for a 1-g/L higher complement C3 concentration was 2.43 (95% CI, 1.74-3.40) (Fig. 4). Corresponding HRs were 2.13 (1.42-3.21) in those with a BMI <30 kg/[m.sup.2] and 1.27 (0.65-2.48) for those with a BMI [greater than or equal to]30 kg/[m.sup.2] (P value for interaction: 3 X [10.sup.-11]). That both these HRs were lower than the overall HR of 2.43 likely is because BMI and complement C3 are highly correlated (Fig. 1). There was no strong evidence of interaction between complement C3 and risk of venous thromboembolism for other common, important risk factors for venous thromboembolism (Fig. 4).
The principal finding in this study is that high concentrations of complement C3 are associated with high risk of venous thromboembolism in the general population. This finding is novel.
Mechanistically, it seems plausible that high concentrations of complement C3 could lead to high risk of venous thromboembolism. Other studies have demonstrated that the complement system is activated during blood clotting, suggesting its participation in thrombus formation (24) and the onset of vascular disease (12). In accordance with this, there is supporting evidence for a procoagulant effect of the complement system, which suggests that the coagulation cascade can be activated either directly or indirectly by complement (25). For example, the osmotic lysis of platelets and endothelial cells caused by the membrane attack complex causes subsequent microparticle release to the bloodstream (26, 27), and these microparticles are potent inducers of thrombin generation and clotting (28). Furthermore, complement activation may potentiate platelet function (29, 30), as platelet activation and aggregation are directly induced by the anaphylatoxin complement C3a and its derivatives, which are activation products of complement C3 (31, 32). Finally, components of the coagulation cascade activate the complement system (33-35).
Coagulation and complement activation are obviously 2 distinct systems with unique pathophysiological roles. Nevertheless, these networks have several common characteristics, which allow for multiple instances of cross-talk and can explain the association ofboth systems with several clinical inflammatory and thrombotic conditions (25). Furthermore, therapeutic complement C3 inhibition prevents complement activation and reduces markers of immune cell activation, inflammation, and coagulation (36). In the alternative pathway of complement activation, spontaneous conversion of complement C3 into C3a and C3b triggers the complement cascade (37). Subsequent steps of the complement cascade lead to formation of several effector molecules, including opsonization of pathogens, formation of the membrane attack complex inducing osmotic lysis of cells, and recruitment and activation of inflammatory cells by the anaphylatoxins complement C3a and C5a (25). Thus, lifelong increased concentrations of complement C3 may biologically be considered a procoagulant state, as also suggested by the present results.
Our finding that BMI attenuated the association between complement C3 and venous thromboembolism requires further discussion. Because obesity is known to lead to venous thromboembolism (23), we therefore speculate that this might in part be through the activation of complement. Such a scenario is likely, since adipose tissue synthesizes complement C3 (19), and since concentrations of complement C3 vary with changing weight (21).
If not causally linked, being an acute-phase reactant, complement C3 could possibly just be a plasma marker of low-grade inflammation in the same way as CRP; however, that the association of high complement C3 concentration and high risk of venous thromboembolism seemed robust even after adjustment for CRP indicates that the observed association is not likely due to systemic inflammation per se.
Major strengths of our study include the large cohort of the general population including 80 517 individuals, the comprehensive survey and information on diagnoses of venous thromboembolism, the large number of events, and the long observation period without loss to follow-up of even a single individual. Moreover, the association between high complement C3 concentrations and high risk of venous thromboembolism was consistent in all analyzed strata, and the only observed interaction was with BMI. Indeed, the fact that the association between complement C3 and venous thromboembolism was most pronounced in those with a BMI of <30 kg/[m.sup.2] supports the notion that this association is not solely driven by obesity.
Potential study limitations include diagnostic misclassification, since some cases of deep venous thrombosis and pulmonary embolism may not have been recorded at hospitals, and some pulmonary embolism events are not a result of venous thromboembolism but rather an emboli made of tumor tissue, fat tissue, or air. Our method of endpoint diagnoses was, however, validated in 2010 with a positive predicted value of 59% for venous thromboembolism diagnosed in any setting and 75% for venous thromboembolism diagnosed during hospital admission (17). However, misclassification ofvenous thromboembolism is likely to be nondifferential, thereby leading only to bias toward the null hypothesis, and therefore cannot explain the present findings. It is also a potential limitation that we studied whites of Danish descent only, although we are not aware of data to suggest that the present results should not apply to other ethnicities and countries. We adjusted for cancer before and after an event, as it is a known risk factor for development of venous thromboembolism, but such adjustment that uses events after baseline may lead to some survivor or diagnostic bias; however, in sensitivity analyses, not adjusting for such cancer events showed similar results. Finally, we had only 1 baseline measurement of complement C3, and since it is an acute-phase reactant, our measurement may not be an expression of the general concentration of complement C3 in each individual. That being said, the concentration of complement C3 in healthy individuals is quite stable over time, with an intraindividual variation of 5% CV (38, 39). Also, variation in complement C3 concentration over time will mean only that the present effect sizes should be viewed as minimal estimations because of regression dilution bias (40).
In conclusion, high concentrations of complement C3 in individuals from the general population were associated with high risk of venous thromboembolism. The association was present even after adjustment for CRP and BMI.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contribution to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors' Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest:
Employment or Leadership: None declared.
Consultant or Advisory Role: None declared.
Stock Ownership: None declared.
Honoraria: None declared.
Research Funding: B. Nordestgaard, Herlev and Gentofte Hospital, Copenhagen University Hospital.
Expert Testimony: None declared.
Patents: None declared.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
Acknowledgments: We thank participants and staff of the Copenhagen General Population Study for their important contributions.
(1.) Naess IA, Christiansen SC, Romundstad P, Cannegieter SC, Rosendaal FR, Hammerstrom J. Incidence and mortality of venous thrombosis: a population-based study. J Thromb Haemost 2007;5:692-9.
(2.) Sogaard KK, Schmidt M, Pedersen L, Horvath-Puho E, Sorensen HT. 30-year mortality after venous thromboembolism: a population-based cohort study. Circulation 2014;130:829-36.
(3.) Ridker PM, Goldhaber SZ, Danielson E, Rosenberg Y, Eby CS, Deitcher SR et al. Long-term, low-intensity warfarin therapy for the prevention of recurrent venous thromboembolism. N Engl J Med 2003;348:1425-34.
(4.) Goon PK, Lip GY. Arterial disease and venous thromboembolism: a modern paradigm? Thromb Haemost 2006;96:111-2.
(5.) Piazza G, Ridker PM. Is venous thromboembolism a chronic inflammatory disease? Clin Chem 2015;61: 313-6.
(6.) Prandoni P, Bilora F, Marchiori A, Bernardi E, Petrobelli F, Lensing AW et al. An association between atherosclerosis and venous thrombosis. N Engl J Med 2003;348: 1435-41.
(7.) Holmqvist ME, Neovius M, Eriksson J, Mantel A, Wallberg-Jonsson S, Jacobsson LT, Askling J. Risk of venous thromboembolism in patients with rheumatoid arthritis and association with disease duration and hospitalization. JAMA 2012;308:1350-6.
(8.) Klovaite J, Nordestgaard BG, Tybjaerg-Hansen A, Benn M. Elevated fibrinogen levels are associated with risk of pulmonary embolism, but not with deep venous thrombosis. Am J Respir Crit Care Med 2013;187:286-93.
(9.) Meyer-Olesen CL, Nielsen SF, Nordestgaard BG. Increased rheumatoid factor and deep venous thrombosis: 2 cohort studies of 54628 individuals from the general population. Clin Chem 2015;61:349-59.
(10.) Reitsma PH, Rosendaal FR. Activation of innate immunity in patients with venous thrombosis: the Leiden Thrombophilia Study. J Thromb Haemost 2004;2: 619-22.
(11.) Zacho J, Tybjaerg-Hansen A, Nordestgaard BG. C-reactive protein and risk of venous thromboembolism in the general population. Arterioscler Thromb Vasc Biol 2010;30:1672.
(12.) Sims PJ, Wiedmer T. The response of human platelets to activated components of the complement system. Immunol Today 1991;12:338-42.
(13.) Engstrom G, Hedblad B, Janzon L, Lindgarde F. Complement C3 and C4 in plasma and incidence of myocardial infarction and stroke: a population-based cohort study. Eur J Cardiovasc Prev Rehabil 2007;14:392-7.
(14.) Muscari A, Bozzoli C, Puddu GM, Sangiorgi Z, Dormi A, Rovinetti Cetal. Association of serum C3 levels with the risk of myocardial infarction. Am J Med 1995;98:357-64.
(15.) Lood C, Eriksson S, Gullstrand B, Jonsen A, Sturfelt G, Truedsson L, Bengtsson AA. Increased C1q, C4 and C3 deposition on platelets in patients with systemic lupus erythematosus-a possible link to venous thrombosis? Lupus2012;21:1423-32.
(16.) Sorensen HT, Horvath-Puho E, Pedersen L, Baron JA, Prandoni P. Venous thromboembolism and subsequent hospitalisation due to acute arterial cardiovascular events: a 20-year cohort study. Lancet 2007;370: 1773-9.
(17.) Severinsen MT, Kristensen SR, Overvad K, Dethlefsen C, Tjonneland A, Johnsen SP. Venous thromboembolism discharge diagnoses in the Danish National Patient Registry should be used with caution. J Clin Epidemiol 2010;63:223-8.
(18.) Gruys E, Toussaint MJ, Niewold TA, Koopmans SJ. Acute phase reaction and acute phase proteins. J Zhejiang Univ Sci B 2005;6:1045-56.
(19.) Choy LN, Rosen BS, Spiegelman BM. Adipsin and an endogenous pathway of complement from adipose cells.J Biol Chem1992;267:12736-41.
(20.) Marubini E, Valsecchi MG. Analysing survival data from clinical trials and observational studies. Chichester: Wiley and Sons; 1995.
(21.) Nestvold TK, Nielsen EW, Ludviksen JK, Fure H, Landsem A, Lappegard KT. Lifestyle changes followed by bariatric surgery lower inflammatory markers and the cardiovascular risk factors c3 and c4. Metab Syndr Relat Disord 2015;13:29-35.
(22.) Ageno W, Becattini C, BrightonT,SelbyR, Kamphuisen PW. Cardiovascular risk factors and venous thromboembolism: a meta-analysis. Circulation 2008; 117:93-102.
(23.) Klovaite J, Benn M, Nordestgaard BG. Obesity as a causal risk factor for deep venous thrombosis: a Mendelian randomization study. J Intern Med 2015;277: 573-84.
(24.) Fischetti F, Durigutto P, Pellis V, Debeus A, Macor P, Bulla R et al. Thrombus formation induced by antibodies to beta2-glycoprotein I is complement dependent and requires a priming factor. Blood 2005;106:2340-6.
(25.) Oikonomopoulou K, Ricklin D, Ward PA, Lambris JD. Interactions between coagulation and complement-their role in inflammation. Semin Immunopathol 2012;34:151-65.
(26.) Hamilton KK, Hattori R, Esmon CT, Sims PJ. Complement proteins C5b-9 induce vesiculation of the endothelial plasma membrane and expose catalytic surface for assembly of the prothrombinase enzyme complex. J Biol Chem 1990;265:3809-14.
(27.) Sims PJ, Faioni EM, Wiedmer T, Shattil SJ. Comple ment proteins C5b-9 cause release of membrane vesicles from the platelet surface that are enriched in the membrane receptor for coagulation factor Va and express prothrombinase activity. J Biol Chem 1988;263: 18205-12.
(28.) Wiedmer T, Esmon CT, Sims PJ. Complement proteins C5b-9 stimulate procoagulant activity through platelet prothrombinase. Blood 1986;68:875-80. (29.) Polley MJ, Nachman RL. Human complement in thrombin-mediated platelet function: uptake of the C5b-9 complex. J Exp Med 1979;150:633-45.
(30.) Zimmerman TS, Arroyove CM, Muller-Eberhard HJ. A blood coagulation abnormality in rabbits deficient in thesixth component of complement (C6)and its correction by purified C6. J Exp Med 1971;134:1591-600.
(31.) Polley MJ, Nachman RL. Human platelet activation by C3a and C3a des-arg. J Exp Med 1983;158:603-15.
(32.) Gushiken FC, Han H, Li J, Rumbaut RE, Afshar-Kharghan V. Abnormal platelet function in C3-deficient mice. J Thromb Haemost 2009;7:865-70.
(33.) Huber-Lang M, Sarma JV, Zetoune FS, Rittirsch D, Neff TA, McGuire SR et al. Generation of C5a in the absence of C3: a new complement activation pathway. Nat Med 2006;12:682-7.
(34.) Amara U, Flierl MA, Rittirsch D, KlosA, Chen H, Acker B et al. Molecular intercommunication between the complement and coagulation systems. J Immunol 2010; 185:5628-36.
(35.) Del C, I, Cruz MA, Zhang H, Lopez JA, Afshar-Kharghan V. Platelet activation leads to activation and propagation of the complement system. J Exp Med 2005;201: 871-9.
(36.) DeAngelis RA, Reis ES, Ricklin D, Lambris JD. Targeted complement inhibition as a promising strategy for preventing inflammatory complications in hemodialysis. Immunobiology 2012;217:1097-105.
(37.) Walport MJ. Complement. First of two parts. N Engl J Med 2001;344:1058-66.
(38.) Butts WC, James GE, Keuhneman M. Intra-individual variation in the concentrations of IgA, IgG, IgM, and complement component C3 in serum of a normal adult population. Clin Chem 1977;23:511-4.
(39.) Sebastian-Gambaro MA, Liron-Hernandez FJ, Fuentes-Arderiu X. Intra-and inter-individual biological variability data bank. Eur J Clin Chem Clin Biochem 1997;35: 845-52.
(40.) Clarke R, Shipley M, Lewington S, Youngman L, Collins R, Marmot M, Peto R. Underestimation of risk associations due to regression dilution in long-term follow-up of pro-spective studies. Am J Epidemiol 1999;150:341-53.
Ina Norgaard, [1,2] Sune F. Nielsen, [1,2] and Borge G. Nordestgaard [1,2] *
 Department of Clinical Biochemistry and the Copenhagen General Population Study, Herlev and Gentofte Hospital, Copenhagen University Hospital, Herlev, Denmark;  FacultyofHealth and Medical Sciences, University of Copenhagen, Copenhagen, Denmark.
* Address correspondence to this author at: Department of Clinical Biochemistry, 54M1, Herlev and Gentofte Hospital, Copenhagen University Hospital, Herlev Ringvej 75, DK-2730, Herlev, Denmark. E-mail firstname.lastname@example.org.
Received October 27,2015; accepted December 16,2015.
Previously published online at DOI: 10.1373/clinchem.2015.251314
 Nonstandard abbreviations: ICD, International Classification of Diseases; BMI, body mass index; CRP, C-reactive protein; HR, hazard ratio.
Caption: Fig. 1. Association between plasma concentrations of complement C3 and BMI and concentrations of CRP. P-values are for Kruskal-Wallis test.
On the basis of 80 517 individuals from the Copenhagen General Population Study.
Caption: Fig. 2. Cumulative incidence of venous thromboembolism, deep venous thrombosis, and pulmonary embolism by tertiles of complement C3 concentrations as a function of age.
Calculated with use of competing risk regression, nonparametric cumulative incidence estimation. On the basis of 80 517 individuals from the Copenhagen General Population Study.
Caption: Fig. 3. Risk of venous thromboembolism, deep venous thrombosis, and pulmonary embolism by tertiles of complement C3 concentrations.
On the basis of 80 517 individuals from the Copenhagen General Population Study.
Caption: Fig. 4. Risk of venous thromboembolism associated with complement C3 on a continuous scale for a 1-g/L higher concentration overall, and in different strata.
On the basis of 80 517 individuals from the Copenhagen General Population Study.
Table 1. Baseline characteristics of individuals from the Copenhagen General Population Study. (a) Plasma complement C3 tertile, g/L Characteristic <1.04 1.04-1.2 n 28 082 25 928 Age, years 57 (48-66) 59 (50-68) Sex Women 15 765 (56) 13 535 (52) Men 12 317 (44) 12 393 (48) BMI, kg/[m.sup.2] 24 (22-26) 26 (24-28) Current smoker 5642 (20) 5276 (20) Cumulative pack-years (c) 18 (4-50) 25 (6-61) Daily cigarettes (or equivalent) smoked (d) 21 (10-35) 26 (14-40) Alkohol units perweek (e) 10 (5-17) 9 (4-16) Use of oral contraceptive (f) 526 (9) 538 (15) Use of hormone replacement therapy (g) 1792 (18) 1656 (16) Leisure-time physical activity per week <2 h exercise 1041 (4) 1455 (6) 2-4 h light exercise 10 406 (37) 11 271 (43) 2-4 h demanding exercise 14 250 (51) 11 643 (45) >4 h exercise 2398 (8) 1569 (6) Marital status Single 2170 (8) 1785 (7) Married 21 334 (76) 19 425 (75) Separated or divorced 2717 (10) 2606 (10) Widowed 1861 (6) 2112 (8) Education Elementary (1-9 years) 5768 (21) 7248 (28) High school (10-12 years) 15 566 (55) 13 965 (54) Academic (>12 years) 6748 (24) 4715 (18) Systolic blood pressure, mmHg 132 (120-147) 138 (125-152) Total cholesterol mg/dL 208 (185-232) 220 (193-247) mmol/L 5.4 (4.8-6.0) 5.7 (5.0-6.4) LDL cholesterol mg/dL 116 (97-139) 127 (104-151) mmol/L 3.0 (2.5-3.6) 3.3 (2.7-3.9) HDL cholesterol mg/dL 66 (54-81) 62 (50-73) mmol/L 1.7 (1.4-2.1) 1.6 (1.3-1.9 CRP, mg/L 1.1 (0.7-1.5) 1.4 (1.0-2.1) Coagulation factors II + VII + X, % 91 (79-105) 98 (85-112) International normalized ratio 1.00 (0.97-1.10) 1.00 (0.90-1.00) Fibrinogen mg/dL 330 (292-374) 367 (323-418) [micro]mol/L 9.7 (8.6-11.0) 10.8 (9.5-12.3) Glucose mg/dL 92 (85-101) 92 (86-103) mmol/L 5.1 (4.7-5.6) 5.1 (4.8-5.7) Any cancer 1 year before or 31 (0.1) 24 (0.1) after an event Hospitalization within 3 948 (3) 983 (4) months before an event Major surgery within 3 months 1 (0.00) 2 (0.01) before an event Rheumatoid factor, IU/mL (h) 13 (10-20) 13 (10-21) Factor V Leiden (R506Q) 1439 (9) 1361 (8) hetero- or homozygous (i) Prothrombin G20210A hetero- or 326 (2) 336 (2) homozygous (i) Plasma complement C3 P for tertile, g/L trend (b) Characteristic >1.2 n 26 507 Age, years 60 (51-68) 4 x [10.sup.-90] Sex 0.14 Women 15 068 (57) Men 11 439 (43) BMI, kg/[m.sup.2] 29 (26-31) <1 x [10.sup.-300] Current smoker 5263 (20) 0.31 Cumulative pack-years (c) 31 (8-68) 7 x [10.sup.-161] Daily cigarettes (or equivalent) smoked (d) 28 (15-40) 2 x [10.sup.-61] Alkohol units perweek (e) 8 (3-15) 2 x [10.sup.-111] Use of oral contraceptive (f) 893 (29) 2 x [10.sup.-29] Use of hormone replacement therapy (g) 1732 (14) 4 x [10.sup.13] Leisure-time physical activity per week <1 x [10.sup.-300] <2 h exercise 2659 (10) 2-4 h light exercise 13 472 (51) 2-4 h demanding exercise 9351 (35) >4 h exercise 1040 (4) Marital status 4 x [10.sup.-38] Single 2055 (8) Married 18 948 (71) Separated or divorced 2977 (1 1) Widowed 2527 (10) Education <1 x [10.sup.-300] Elementary (1-9 years) 9373 (35) High school (10-12 years) 13 268 (50) Academic (>12 years) 3866 (15) Systolic blood pressure, mmHg 142 (130-157) <1 x [10.sup.-300] Total cholesterol <1 x [10.sup.-300] mg/dL 228 (197-255) mmol/L 5.9 (5.1-6.6) LDL cholesterol <1 x [10.sup.-300] mg/dL 135 (108-162) mmol/L 3.5 (2.8-4.2) HDL cholesterol <1 x [10.sup.-300] mg/dL 54 (42-69) mmol/L 1.4 (1.1-1.8) CRP, mg/L 2.2 (1.4-4.1) <1 x [10.sup.-300] Coagulation factors II + VII + X, % 105 (91-119) <1 x [10.sup.-300] International normalized ratio 0.91 (0.90-1.00) <1 x [10.sup.-300] Fibrinogen <1 x [10.sup.-300] mg/dL 411 (360-476) [micro]mol/L 12.1 (10.6-14.0) Glucose 7 x [10.sup.-115] mg/dL 94 (86-104) mmol/L 5.2 (4.8-5.8) Any cancer 1 year before or 40 (0.2) 0.18 after an event Hospitalization within 3 1185 (4) 4 x [10.sup.-11] months before an event Major surgery within 3 months 2 (0.01) 0.55 before an event Rheumatoid factor, IU/mL (h) 13 (10-23) 0.32 Factor V Leiden (R506Q) 1507 (9) 0.55 hetero- or homozygous (i) Prothrombin G20210A hetero- or 388 (2) 0.19 homozygous (i) (a) Data are n (%)or median (interquartile range). (b) Stata function nptrend. (c) Among current and former smokers. (d) Among currentsmokers. (e) 1 unit alcohol = 12 g. (f) Among premenopausal women. (g) Among postmenopausal women. (h) Measured in 43 944 individuals. (i) Genotyped in 51 025 individuals. Table 2. Associations and correlations among complement C3 and primary predictors of venous thromboembolism in 80 517 individuals from the Copenhagen General Population Study. Correlation Complement Association C3 Sex Complement C3 [R.sup.2] = -0.03 Sex P < 0.0001 Age P < 0.0001 P < 0.0001 BMI P < 0.0001 P < 0.0001 Smoking status P < 0.0001 P < 0.0001 CRP P < 0.0001 P = 0.80 Factor V Leiden (a) P = 0.86 P = 0.96 Prothrombin P = 0.39 P = 0.52 G20210A (a) Correlation Association Age BMI Complement C3 [R.sup.2] = 0.05 [R.sup.2] = 0.53 Sex [R.sup.2] = 0.03 [R.sup.2] = 0.14 Age [R.sup.2] = 0.06 BMI P < 0.0001 Smoking status P < 0.0001 P < 0.0001 CRP P < 0.0001 P < 0.0001 Factor V Leiden (a) P = 0.92 P = 0.22 Prothrombin P = 0.04 P = 0.46 G20210A (a) Correlation Smoking Association status CRP Complement C3 [R.sup.2] = 0.03 [R.sup.2] = 0.29 Sex [R.sup.2] = 0.07 [R.sup.2] = -0.0009 Age [R.sup.2] = 0.11 [R.sup.2] = 0.09 BMI [R.sup.2] = 0.05 [R.sup.2] = 0.15 Smoking status [R.sup.2] = 0.03 CRP P < 0.0001 Factor V Leiden (a) P = 0.12 P = 0.16 Prothrombin P = 0.60 P = 0.20 G20210A (a) Correlation Factor V Prothrombin Association Leiden G20210A Complement C3 [R.sup.2] = 0.0008 [R.sup.2] = 0.004 Sex [R.sup.2] = -0.0002 [R.sup.2] = -0.003 Age [R.sup.2] = 0.004 [R.sup.2] = -0.009 BMI [R.sup.2] = -0.005 [R.sup.2] = 0.003 Smoking status [R.sup.2] = -0.007 [R.sup.2] = -0.002 CRP [R.sup.2] = -0.006 [R.sup.2] = 0.006 Factor V Leiden (a) [R.sup.2] = -0.005 Prothrombin P = 0.24 G20210A (a) (a) For factor V Leiden R506Q (rs6025) and prothrombin G20210A (rs1799963), data were from 51 025 individuals only.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Hemostasis and Thrombosis|
|Author:||Norgaard, Ina; Nielsen, Sune F.; Nordestgaard, Borge G.|
|Date:||Mar 1, 2016|
|Previous Article:||Mass spectrometry-based adrenal and peripheral venous steroid profiling for subtyping primary aldosteronism.|
|Next Article:||Undetectable alanine aminotransferase during hospitalization.|