Smoking
2016; Lippincott Williams & Wilkins; Volume: 135; Issue: 1 Linguagem: Inglês
10.1161/circulationaha.116.025024
ISSN1524-4539
Autores Tópico(s)Cardiovascular Health and Disease Prevention
ResumoHomeCirculationVol. 135, No. 1Smoking Free AccessEditorialPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessEditorialPDF/EPUBSmokingA Dual Pathogen for Arterial and Venous Thrombosis Sonia S. Anand, MD, PhD Sonia S. AnandSonia S. Anand From Departments of Medicine and Epidemiology, Population Health Research Institute, McMaster University, Hamilton, ON, Canada. Originally published3 Jan 2017https://doi.org/10.1161/CIRCULATIONAHA.116.025024Circulation. 2017;135:17–20Article, see p 7Traditional book learning and clinical training of common risk factors for venous thromboembolism (VTE) did not include cigarette smoking. I recall as a medical student, after giving the history of a patient presenting with deep venous thrombosis, being told that smoking was not a risk factor for venous disease, only arterial disease. This perplexed me. If smoking was known to "thicken the blood," as evidenced by the increased fibrinogen, and venous thromboses typically represent a larger clot burden compared with arterial thromboses, why wouldn't cigarette smoking be a dual pathogen for venous and arterial disease?Prior evaluations have compared the risk of smoking and other traditional arterial risk factors (ie, diabetes mellitus, hypertension, and elevated cholesterol) for arterial vascular disease with the risk of these factors for VTE.1 Our understanding of the pathophysiology of arterial vascular diseases is based on atherosclerosis, which becomes atherothrombosis when plaque rupture occurs, platelets are recruited, and the coagulation cascade is activated. Clinically apparent venous thrombosis is usually the result of venous stasis (immobility or obstruction), stimulation of the coagulation cascade from circulating or exogenous inciters leading to a hypercoagulable state, or mechanical injury to the vessel wall. One may expect that there are unique risk factors for arterial and venous thrombosis and some shared risk factors for which stasis conditions, hypercoagulable states, and vessel injury are the inciters (Figure).Download figureDownload PowerPointFigure. Risk factors for arterial and venous thrombosis. HRT indicates hormone replacement therapy; LV, left ventricle; MI, myocardial infarction; and OCP, oral contraceptive pills.Some evidence supports a shared pathogenesis of arterial and venous thrombosis. The Tromsø study of 21 624 subjects recruited from Norway showed that a family history of myocardial infarction is associated with a 52% increased risk of myocardial infarction and a 26% increased risk of VTE after adjustment for traditional risk factors for VTE.2 Furthermore, both hormone replacement therapy and oral contraceptive pills are associated with an increased risk of myocardial infarction, stroke, and VTE, and their effects are magnified in the presence of other thrombosis promoters such as genetic polymorphisms (ie, factor V Leiden) and cigarette smoke.Comparison studies of arterial and venous risk factors have produced mixed results, in large part because cardiovascular disease outcomes are 2.5 to 3.5 times more common in similar cohorts of middle-aged adults who are apparently healthy.3,4 Prior analyses from the US Physicians Health Study reported no association between hypertension, elevated cholesterol, diabetes mellitus, and smoking with VTE, whereas increasing body mass index (BMI) and height were associated with incident VTE more than with cardiovascular disease.3 Notably, only 358 VTE events were observed in this study compared with 2300 cardiovascular disease events, which makes the null finding of smoking as a risk factor for VTE potentially a type 2 error. Furthermore, although meta-analysis may overcome the relatively low number of VTE events observed in individual cohorts, the correlation between putative risk factors and heterogeneity in study design, including definitions of risk factors, may be more difficult to overcome and may give misleading signals. This was shown in the 2008 meta-analysis by Ageno et al1 in which hypertension, hyperlipidemia, and diabetes mellitus were reported to be associated with VTE but smoking was not. However, this meta-analysis used published data, and no adjustments for age or BMI were made.The study published by Mahmoodi et al5 in this issue of Circulation represents an advance over prior studies because it is an individual-level meta-analysis (9 of potential 13 prospective studies provided data) involving 244 865 participants and is well powered (n=4910 VTE events), with harmonized risk factor definitions across studies, and careful measurement of the outcomes of interest. The outcomes of VTE used strict criteria, including duplex ultrasound or venography for deep venous thrombosis and ventilation-perfusion scan, pulmonary angiography, computed tomography pulmonary angiograms, or autopsy for pulmonary embolism, overcoming limitations of prior studies in which self-report or physician report was used. The same definition of risk factors was used across studies for hypertension (ie, elevated measured blood pressure [BP] or use of antihypertensive medications), diabetes mellitus (ie, elevated glucose using preset criteria, use of glucose-lowering medications, or self-reported diabetes mellitus), and hyperlipidemia (ie, elevated total cholesterol or use of lipid-lowering medications), and smoking was self-reported and classified as current versus former/never and former versus never smokers. No dose response in terms of light versus heavy smokers was presented. BMI was measured and used as a continuous variable. However, the classification of provoked versus unprovoked VTE varied across studies, and although the majority included invasive surgery and cancer as provoking factors, the time of exposure to the provoking factor before the VTE diagnosis, and the classification of exposure to exogenous hormones as a provoking factor were inconsistent.In this meta-analysis, not only does current smoking emerge as a risk factor for VTE, but no association between hypertension, hyperlipidemia, or diabetes mellitus and VTE was identified after adjustment for confounders. The hazard ratio (HR) of smoking with VTE is 1.19 (95% confidence interval [CI], 1.08–1.32) after adjustment for age, sex, and BMI. Former smoking was not significantly associated with VTE (HR, 0.99; 95% CI, 0.93–1.06). It is informative to review that before adjustment for BMI the association between smoking and VTE was not apparent, but because smoking is correlated with lower BMI, failure to adjust for BMI leads to a confounded estimate. This demonstrates the impact of performing a high-powered (by number of outcome events) meta-analysis and being able to adjust for covariates such as BMI with the use of individual participant data. The authors observed the association between current smoking and VTE (HR, 1.36; 95% CI, 1.22–1.52) for which there was an already identified cause (so-called provoked VTE) and no association between current smoking and VTE for which there was no known cause (so-called unprovoked VTE; HR, 1.08; 95% CI, 0.90–1.29). However, as a result of possible misclassification of provoked versus unprovoked (eg, in the WHI [Women's Health Initiative], use of hormonal therapy [active arms] was not considered for classifying VTE as provoked), the overall HR of current smoking and VTE is the most robust. As the authors suggest, the association of current smoking with provoked VTE could be enhanced by comorbid conditions associated with smoking such as cancer or other smoking-related hospitalizations.The authors went on to assess the continuous measures of risk factors, including systolic and diastolic BPs, cholesterol, and glucose, using cubic spline analysis. Although hypertension itself was not a risk factor for VTE, the continuous measurement of systolic BP (160 versus 110 mm Hg) was inversely associated with VTE (0.79; 95% CI, 0.68–0.92).In observational studies, it is difficult to prove absolute causality because of potential threats to validity. Often, the Bradford Hill criteria,6 including strength of association (effect size), consistency across studies, biological plausibility, temporality, dose-response relationship, coherence, and experimental evidence, are invoked to help inform us of the strength of the evidence. Applying these criteria, Mahmoodi et al demonstrate in their meta-analysis that only some of the criteria are met for smoking and VTE, and even fewer are met for the association between the continuous measure of BP and VTE.The initial HR of smoking with VTE in the unadjusted model was nonsignificant (0.87; 95% CI, 0.72–1.05), whereas the association with current smoking becomes apparent after adjustment for age, sex, and BMI because smoking is associated with lower BMI. BMI fits the definition of a confounder, a variable that is associated with both the risk factor and the outcome but is not part of the causal pathway between the exposure and the outcome. The effect size of the association of smoking with VTE is low yet appears to be generally consistent across the large individual studies that contribute to the meta-analysis (ARIC [Atherosclerosis Risk in Communities], DCH [Danish Diet, Cancer and Health], Tromsø, WHI). There is biological plausibility because smoking exposure is associated with increased markers of hypercoagulability, the smoking status is ascertained at baseline, and VTE events are incident. We cannot assess the data set for a dose-response relationship because the data by dose of smoking (ie, number of cigarettes per day) are not provided, although the HR of former smoking is lower than for current smoking, which makes biological sense and is supported by the finding of a previous meta-analysis of smoking and VTE risk that demonstrated a dose-response relationship.7 Furthermore, there are no experimental data for, that is, randomized trial evidence of, a lower incidence of VTE in those who achieved smoking cessation. On the other hand, the inverse association observed with the continuous measure of systolic BP (160 mm Hg versus the reference value of 110 mm Hg; relative risk, 0.70; 95% CI, 0.68–0.92) was associated with a lower incidence of mainly unprovoked VTE, which may be postulated to exist as a result of various biological mechanisms (but they would all be speculative), could have occurred by chance, or alternatively could be the result of correlation with another condition (ie, comorbid conditions such as atrial fibrillation for which oral anticoagulants are used) or confounding. Although the authors took care to adjust for some known covariables, they did not adjust for all covariables such as estrogen/progesterone use in WHI, which is the largest contributor of patients and outcomes in this meta-analysis. The association between higher systolic BP and a lower risk of VTE is inconsistent in the literature. Hypertension itself is not associated with VTE after adjustment for covariates, and the only other large assessment of BP lowering suggested the opposite association, with a reduction of systolic BP of 20 mm Hg being associated with a reduced risk of pulmonary embolism.8 This lack of coherency across analyses suggests that this result may have occurred by chance or is confounded.Although elevated cholesterol is not associated with VTE in this meta-analysis, prior clinical studies in which patients were treated with statins have reported that statin use is associated with a reduced incidence of VTE.9,10 However, the point estimate in the Mahmoodi et al analysis of elevated cholesterol and VTE in the fully adjusted model is 0.96 (95% CI, 0.87–1.05), and although statin use may have biased the elevated cholesterol effect size toward the null, we cannot assess this because the authors do not stratify the data by statin use versus no statin use. It is possible that the other effects of statin beyond cholesterol lowering such as reduction of inflammation may play a role in the apparent reduction of incident VTE, which future studies should address.To summarize, we learn a great deal from the meta-analysis by Mahmoodi and colleagues, for example, that smokers have a higher risk of VTE, whereas there are no strong signals for hypertension, diabetes mellitus, and elevated cholesterol. Although the effect size is low, it is likely that many cases of VTE are caused by the presence of multiple risk factors, all with small effects (ie, risk ratios <2.0) that increase risk above a certain threshold and clinical thrombosis develops. It also demonstrates that caution must be given to small studies and meta-analyses of small studies in which errors can be multiplied and not obviated. The best way forward to characterize risk factors for relatively uncommon conditions is to design and conduct large studies or to conduct individual-level patient meta-analyses with careful consideration of and adjustment for potential confounders.DisclosuresNone.FootnotesThe opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.Circulation is available at http://circ.ahajournals.org.Correspondence to: Sonia S. Anand, MD, PhD, Department of Medicine & Epidemiology, McMaster University, Population Health Research Institute, Hamilton Health Sciences and McMaster University, 1280 Main St W, MDCL 3204, Hamilton, ON L8S 4K1, Canada. E-mail [email protected]References1. Ageno W, Becattini C, Brighton T, Selby R, Kamphuisen PW. Cardiovascular risk factors and venous thromboembolism: a meta-analysis.Circulation. 2008; 117:93–102. doi: 10.1161/CIRCULATIONAHA.107.709204.LinkGoogle Scholar2. Lind C, Enga KF, Mathiesen EB, Njølstad I, Brækkan SK, Hansen JB. Family history of myocardial infarction and cause-specific risk of myocardial infarction and venous thromboembolism: the Tromsø Study.Circ Cardiovasc Genet. 2014; 7:684–691. doi: 10.1161/CIRCGENETICS.114.000621.LinkGoogle Scholar3. Glynn RJ, Rosner B. Comparison of risk factors for the competing risks of coronary heart disease, stroke, and venous thromboembolism.Am J Epidemiol. 2005; 162:975–982. doi: 10.1093/aje/kwi309.CrossrefMedlineGoogle Scholar4. Tsai AW, Cushman M, Rosamond WD, Heckbert SR, Polak JF, Folsom AR. Cardiovascular risk factors and venous thromboembolism incidence: the longitudinal investigation of thromboembolism etiology.Arch Intern Med. 2002; 162:1182–1189.CrossrefMedlineGoogle Scholar5. Mahmoodi BK, Cushman M, Naess IA, Allison MA, Bos WJ, Braekkan SK, Cannegieter SC, Ganswvoort RT, PN Gona, Hammerstrøm J, Hansen JB, Heckbert S, Holst AG, Lakoski SG, Lutsey PL, Manson JE, Martin LW, Matsushita K, Meijer K, Overvad K, Prescott E, Puurunen M, Rossouw JE, Sang Y, Severinsen MT, ten Berg J, Folsom AR, Zakai NA. Association of traditional cardiovascular risk factors with venous thromboembolism: an individual participant data meta-analysis of prospective studies.Circulation. 2017; 135:7–16. doi: 10.1161/CIRCULATIONAHA.116.024507.LinkGoogle Scholar6. Hill AB. The environment and disease: association or causation?Proc R Soc Med. 1965; 58:295–300.CrossrefMedlineGoogle Scholar7. Cheng YJ, Liu ZH, Yao FJ, Zeng WT, Zheng DD, Dong YG, Wu SH. Current and former smoking and risk for venous thromboembolism: a systematic review and meta-analysis.PLoS Med. 2013; 10:e1001515. doi: 10.1371/journal.pmed.1001515.CrossrefMedlineGoogle Scholar8. Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002; 360:1903–1913.CrossrefMedlineGoogle Scholar9. Glynn RJ, Danielson E, Fonseca FA, Genest J, Gotto AM, Kastelein JJ, Koenig W, Libby P, Lorenzatti AJ, MacFadyen JG, Nordestgaard BG, Shepherd J, Willerson JT, Ridker PM. A randomized trial of rosuvastatin in the prevention of venous thromboembolism.N Engl J Med. 2009; 360:1851–1861. doi: 10.1056/NEJMoa0900241.CrossrefMedlineGoogle Scholar10. Lassila R, Jula A, Pitkäniemi J, Haukka J. The association of statin use with reduced incidence of venous thromboembolism: a population-based cohort study.BMJ Open. 2014; 4:e005862. doi: 10.1136/bmjopen-2014-005862.CrossrefMedlineGoogle Scholar Previous Back to top Next FiguresReferencesRelatedDetailsCited By Berkowitz S, Bauersachs R, Szarek M, Nehler M, Debus E, Patel M, Anand S, Capell W, Hess C, Hsia J, Leeper N, Brasil D, Mátyás L, Diaz R, Brodmann M, Muehlhofer E, Haskell L and Bonaca M (2022) Prevention of arterial and venous thrombotic events in symptomatic peripheral arterial disease patients after lower extremity revascularization in the VOYAGER PAD trial: Dual anticoagulant/antiplatelet regimen vs antiplatelet therapy alone, Journal of Thrombosis and Haemostasis, 10.1111/jth.15673, 20:5, (1193-1205), Online publication date: 1-May-2022. Reitz K, Althouse A, Meyer J, Arya S, Goodney P, Shireman P, Hall D and Tzeng E (2022) Association of Smoking With Postprocedural Complications Following Open and Endovascular Interventions for Intermittent Claudication, JAMA Cardiology, 10.1001/jamacardio.2021.3979, 7:1, (45), Online publication date: 1-Jan-2022. Michailidou D, Zhang T, Stamatis P and Ng B (2022) Risk of venous and arterial thromboembolism in patients with giant cell arteritis and/or polymyalgia rheumatica: A Veterans Health Administration population‐based study in the United States, Journal of Internal Medicine, 10.1111/joim.13446, 291:5, (665-675), Online publication date: 1-May-2022. Landzberg D, Nogueira R, Al-Bayati A, Kim S, Bouslama M, Pisani L, da Camara C, Frankel M, Nahab F, Bianchi N and Haussen D (2021) Baseline Characteristics of Patients with Symptomatic Carotid Webs: A Matched Case Control Study, Journal of Stroke and Cerebrovascular Diseases, 10.1016/j.jstrokecerebrovasdis.2021.105823, 30:8, (105823), Online publication date: 1-Aug-2021. Xu Y, Siegal D and Anand S (2020) Ethnoracial variations in venous thrombosis: Implications for management, and a call to action, Journal of Thrombosis and Haemostasis, 10.1111/jth.15140, 19:1, (30-40), Online publication date: 10-Jan-2021. Thomas D (2019) Sevrage tabagique : un défi pour les cardiologues et les pneumologues, Revue des Maladies Respiratoires, 10.1016/j.rmr.2018.03.008, 36:4, (527-537), Online publication date: 1-Apr-2019. Puvanesarajah V, Rao S, Hassanzadeh H and Kebaish K Determinants of perioperative transfusion risk in patients with adult spinal deformity, Journal of Neurosurgery: Spine, 10.3171/2017.10.SPINE17884, 28:4, (429-435) Szylińska A, Kotfis K, Bott-Olejnik M, Wańkowicz P and Rotter I (2022) Post-Stroke Outcomes of Patients with Chronic Obstructive Pulmonary Disease, Brain Sciences, 10.3390/brainsci12010106, 12:1, (106) January 3, 2017Vol 135, Issue 1 Advertisement Article InformationMetrics © 2016 American Heart Association, Inc.https://doi.org/10.1161/CIRCULATIONAHA.116.025024PMID: 28028060 Originally publishedJanuary 3, 2017 KeywordsEditorialsvascular diseasesarterial thrombosisvenous thromboembolismbody mass indexmeta-analysissmokingPDF download Advertisement SubjectsCardiovascular DiseaseEmbolismRisk FactorsThrombosis
Referência(s)