International Journal of Obesity (2003) 27, S29–S34 & 2003 Nature Publishing Group All rights reserved 0307-0565/03 $25.00 www.nature.com/ijo PAPER Emerging relationships of inflammation, cardiovascular disease and chronic diseases of aging
RP Tracy1*
1Laboratory for Clinical Biochemistry Research, Department of Pathology, University of Vermont, Colchester, VT, USA
The last 10 y have seen an enormous surge in research focused on inflammation and atherosclerotic heart disease. In parallel, inflammation (used as a term to represent a broad array of response systems) has become a topic of interest in a number of different areas of chronic disease including type II diabetes, cognitive decline and frailty, among others. These discoveries are opening up many new opportunities for risk assessment. For example, markers of inflammation such as C-reactive protein are becoming established as important additions in helping to define those at the greatest risk of progressive vascular disease. These discoveries are also important in the area of risk management. Older medications now in wide use are being found to have previously unknown anti-inflammatory effects (eg, statins); these effects are now viewed as being important to the overall effectiveness of these compounds. In addition, the many different aspects of inflammation provide a wealth of targets for new therapeutics, which will be increasingly important as the population continues to age. Although daunting in complexity, studies on the relation of inflammation to disease have already proven useful, and hold promise for providing fundamental advances in both basic biology and clinical medicine. International Journal of Obesity (2003) 27, S29–S34. doi:10.1038/sj.ijo.0802497
Keywords: inflammation; cardiovascular disease; elderly
Introduction before, thrombosis had been identified as the proximate Over 150 y ago, Virchow recognized the inflammatory cause of myocardial infarctions (MI), so the fibrinogen component of atherosclerosis, and commented on it in his findings were initially interpreted in the context of fibrino- writings.1 It has just been within the last 10 y, however, that gen’s role as a coagulation factor. However, fibrinogen was a we have rediscovered and clearly identified the important known acute-phase protein, and the possibility that higher association between atherosclerosis and the wide variety of values were a marker of underlying inflammation was not activities that are components of what might be considered completely overlooked. ‘inflammation’. Many areas of research have contributed to The second important observation was that higher plasma this new emphasis, including molecular anatomy and levels of acute-phase proteins such as C-reactive protein histology, biochemistry, cell biology, and molecular biol- (CRP) and serum amyloid A were associated with increased ogy/genetics. Also, and in particular, molecular epidemiol- risk of poor prognosis in those with established cardiovas- ogy has played an important role; this is the area I will cular disease.3 This work extended the fibrinogen findings emphasize in this brief review. and emphasized the inflammatory nature of these markers, From a population biology standpoint, two important since these were all acute-phase proteins. findings preceded this new era, and laid the groundwork for These two findings were followed by a large number of inflammation as an important component of atherosclerosis epidemiological studies in otherwise healthy individuals, and heart disease. The first was the work of Meade et al2 and many of which focused on CRP, demonstrating that even others in the 1980s identifying fibrinogen as a risk marker of within the so-called ‘normal range’, higher values of CRP approximately equal strength to that of cholesterol. Not long compared to lower values indicated increased cardiovascular disease (CVD) risk.4 Throughout this research, however, the essential question remained: did the levels of these plasma *Correspondence: Dr RP Tracy, Professor of Pathology and Biochemistry, proteins simply reflect the underlying pathophysiology, or Senior Associate Dean for Research and Academic Affairs, 208 South Park Drive, Suite 2, University of Vermont, Colchester VT 05446, USA. did these proteins actually participate in the disease process? E-mail: [email protected] A third position existed, which is that both of the above are Inflammation and chronic diseases of old age RP Tracy S30 been described in detail by us and others. Essentially, the major correlates of CRP levels in a healthy population are: (a) the components of the metabolic syndrome, especially adiposity; (b) diabetes status; (c) coagulation activity; and (d) smoking status. Other correlates are weaker, including HDL-cholesterol levels (negative association), sex (women tend to have higher values than men), and measures of atherosclerotic heart disease such as ultrasonographic esti- mates of carotid artery wall thickness and ECG abnormal- ities. In particular, CRP is not strongly associated with the newest measure of subclinical atherosclerosis, coronary calcification. In postmenopausal women, estrogen use is associated with B2-fold higher values, and estrogen has been shown to increase CRP. The first study of CRP and CVD risk in otherwise healthy people was done in the MRFIT cohort as a nested case– Figure 1 Complex interplay of atherothrombotic disease with plasma control study. CRP was an independent risk factor for future markers of CVD risk: a positive feed-back pathway. This figure illustrates three CVD events, especially in MRFIT smokers. This was followed possible ways of looking at the relation of inflammation mediators and markers to atherothrombotic disease. We propose that the third model is most up by nested case–control studies in the Cardiovascular likely, that is, that vascular disease is in fact an inflammatory process, and that Health Study, the Rural Health Promotion Project, and the markers of inflammation reflect this underlying process, but that these markers Physician Health Study. Many studies have now been done also participate in the process, making it worse. This pathophysiological in the US and elsewhere, both nested case–control studies participation is modulated by an individual’s genetic architecture, nonvascular inflammatory burden, and ability to produce proinflammatory responses in and cohort studies, which have yielded consistent results: in cells such as adipocytes and monocytes. middle aged men and women, CRP is a CVD risk factor that is independent of other known CVD risk factors. As was the correct. These possibilities are illustrated in Figure 1. We have case for fibrinogen in the original Northwick Park Heart consistently supported the third position, since there is a Study findings, this inflammation marker is approximately considerable body of work to support each side of the initial as predictive as plasma lipid levels, and in the recent question.5 We will return to this issue later in the review. Women’s Health Study, was actually a stronger predictor Finally, what do we mean by the term inflammation? In that LDL-cholesterol level.7 this context, we are not discussing the process characterized One of the key correlates of CRP is adiposity, especially classically by redness, swelling, heat, and pain. Rather, we are visceral adiposity. Body fat has received renewed attention discussing a type of ‘microinflammation’, possibly localized recently as a CVD risk factor since this tissue has been to discrete spots in the vasculature, which is mediated by the recognized as a key site for proinflammatory cytokine same cytokines and effector molecules as the classic process production.8 It is likely that increased chronic inflammation but differs markedly in degree. Using CRP as an example, is one of the ways that visceral adiposity manifests its prior to the last decade, CRP levels were considered to be proatherosclerotic effects. Visceral fat is also a major site for either ‘normal’ (o10 mg/l) or ‘abnormal’ (410 mg/l). CRP the production of the key fibrinolysis regulator, plasminogen has a large dynamic range (B1000-fold), and most assays in activator inhibitor-1 (PAI-1). This has been proposed as use at that time did not even return a value below 10 mg/l, another mechanism by which body fat imposes athero- since the distribution of values below this level was not thrombotic risk. considered informative. Owing to the sensitive assays Several investigators have suggested that CRP might be developed by Pepys’ lab (a radioimmunoassay) and our lab useful clinically in CVD risk assessment, along with plasma (an ELISA),6 we now know that virtually all of the risk lipid levels. CRP levels are associated with most of the major prediction contained in the CRP distribution is within the components of the Framingham risk score, such as diabetes, ‘normal range’, that is, below 5 mg/l. It is important to note obesity, hypertension, sex, and smoking. Therefore, in some that while the term ‘high sensitivity CRP’, or ‘hs-CRP’, has ways, CRP may be an integrated reflection of the damage come into use for assays with enough sensitivity to measure caused by these components. While CRP may add to the CRP levels o10 mg/l, there is no biochemical difference Framingham score in some populations, it may have only between CRP and hs-CRP. moderate added value in others. In any population, CRP may prove useful in those with moderate risk, where a low CRP value might suggest a conservative approach, and a high CRP might suggest a more aggressive approach. Also, there are CRP as a marker of inflammation difficulties with using the Framingham score, such as its In recent years, studies of inflammation and CVD risk have complexity, which have limited the use of this method. It focused predominantly on CRP. The correlates of CRP have may be time to consider a simple, laboratory-based risk score
International Journal of Obesity Inflammation and chronic diseases of old age RP Tracy S31 (eg, CRP, LDL-, and HDL-cholesterol), which would contain Consistency of association with CVD over time much of the information in the Framingham score, but be The majority of the data that are currently available in the more easily and more consistently applied. literature concerning inflammation and CVD risk assessment refer to middle-aged men and women, as reviewed above. However, there are some data to support the position that inflammation is associated with atherosclerotic disease in both older people and younger people. As part of the Other markers of inflammation longitudinal epidemiology study called the Cardiovascular As mentioned above, a number of plasma molecules have Health Study (CHS; 5888 men and women 464 y at been studied recently, and are proposed as new risk markers. baseline), we reported that both CRP and fibrinogen were A simplified framework is offered in Table 1. All of the significant, independent CVD risk factors in the elderly. We markers mentioned in this table, as well as others, have have also observed this for the elderly Japanese American either been associated with prevalent CVD, used to predict men in the Honolulu Heart Program (HHP). Also in both the future CVD events, or both. Within this framework, CRP is CHS and the HHP, and the Iowa 65 þ Rural Health Study, seen as just one of a number of plasma markers whose levels CRP and fibrinogen predicted CVD- and all-cause mortality. reflect the many aspects of low-level, chronic inflammation. Therefore, it is clear that inflammation markers are asso- While each of these markers may be seen as a reflection of ciated with both fatal and nonfatal CVD events in the underlying inflammation, for each there is also one or elderly. several plausible scenarios which puts it in the causal In the young, there is less information available. Event- pathway. This phenotype discovery work, whereby research based population data are difficult to obtain due to the continues to uncover new plasma markers of the athero- necessarily long follow-up periods. Cross-sectional data, thrombotic disease process, is important because it leads to a however, can be obtained. In otherwise healthy young adults more complete understanding of the relevant mechanisms (18–28 y) entered in the CARDIA study of CVD risk factors, and the discovery of new mechanisms in atherothrombosis. fibrinogen levels were associated with most of the traditional However, these findings also beg the question: which of CVD risk factors, suggesting a pattern that begins at early these markers is most useful in clinical CVD risk assessment? age. In similar aged individuals with type I diabetes, CRP HDL-cholesterol, while consistently negatively associated levels were correlated with both mean and maximum carotid with inflammation, probably plays a bigger role in lipid wall thickness; results were similar although less strong in metabolism. Fibrinogen and CRP have both been used in nondiabetic controls. In 19–39 y-old subjects with child- many studies, and yield about the same level of informa- hood-onset chronic renal failure, compared to controls, 4 tion. The others have been studied to a lesser degree. Of CRP there was considerable evidence of atherosclerotic disease and fibrinogen, we have argued that CRP is probably the such as increased carotid wall thickness and coronary superior analyte based on technical reasons such as standar- calcification. CRP was strongly correlated to the degree of 5,9 dization, ease of the assay, reproducibility, etc. arterial disease in these subjects. In even younger subjects, increased inflammatory marker levels have been associated with obesity and various indices 10 Table 1 Simplified framework for plasma risk markers related to inflamma- of fat deposition in children, while lower values have been tion, with examples associated with higher physical activity status and increased 11 Systemic markers: hepatic acute phase proteins exercise. In a group of young people, average age 10 y, CRP CRP levels were shown to be strongly associated with endothelial Fibrinogen function as estimated by flow-mediated brachial artery Serum amyloid A dilatation.12 Ceruloplasmin (copper transport) Transferrin (iron transport) Albumin (negative) HDL-cholesterol (apo A-I) Association with different pathology over time Systemic markers: coagulation activation Fibrinopeptides (D-dimer) The nature of the underlying CVD in people changes with Plasmin generation (plasmin-antiplasmin complex) age, as does the nature of the association of inflammation with CVD events. For example, in white men dying sudden Systemic/localized mediators: cytokines cardiac deaths in their 4th and 5th decades, 50–70% present Interleukin-6 TNF-a at autopsy with acute thrombosis, while this number drops to less than a third by the 7th decade, with a much greater Localized mediators: adhesion proteins, enzymes percentage being stable plaque-based coronary deaths.13 This ICAM-1 change has implications for the use of a risk marker such as E-selectin Matrix metalloproteinases CRP in different aged populations. In younger and/or healthier people, where plaque rupture and thrombosis
International Journal of Obesity Inflammation and chronic diseases of old age RP Tracy S32 predominate as the causes of fatal coronary events, CRP has levels are a major risk factor,17 as originally proposed; and long-term prediction. For example, in the Physicians Health results for factor VIII are probably best interpreted as Study, elevated baseline CRP predicted MI equally strongly in supporting inflammation rather than hypercoagulability. D- the first 2 y follow-up period (0–2 y), in the next 2 y period dimer is an integrated marker of coagulant and subsequent (3–4 y), and so on out to 8 y.14 We identified a similar finding fibrinolytic activity; levels do not predict early calcification in in the Honolulu Heart Program cohort, with up to 15 y of younger people (Jacobs, manuscript in preparation), and were follow-up. However, when used to predict CVD mortality in not associated with incident events in a healthy middle-aged the elderly, where stable-plaque-mediated coronary deaths population such as the Physician’s health Study.18 As the predominate, CRP and fibrinogen exhibited a marked population under study has greater atherosclerotic burden, interaction with time-to-event from baseline; that is, these this marker does start to predict events: D-dimer levels were markers were much stronger predictors of fatal CVD deaths predictive in a study of men with peripheral vascular disease, when the event occurred closer to the baseline (N Jenny, and in studies of older middle-aged men with mixed health manuscript in preparation). We have termed this ‘proximate status.19 In the elderly cohort of CHS, D-dimer, as well as the pathophysiology’ to indicate the sensitivity of inflammatory marker of plasmin activation, Plasmin-AntiPlasmin Complex, markers to changes in the underlying disease process that were strongly predictive of events.20 A meta-analysis demon- portend soon-to-occur major events. strated moderate risk prediction for D-dimer.19 The best While it would be interesting to examine the relationship interpretation of these data is that D-dimer reflects the degree of risk markers to atherosclerosis in the very young, this is of atherosclerosis and vascular damage present, not a pre- not yet possible in humans. However, the mouse offers a existing hypercoagulable state. This is supported by the model of early atherosclerosis, and inflammation is clearly findings of Lowe et al,21 where D-dimer predicted future involved in the earliest stages. In a set of experiments in C57/ ischemic events, but other markers of procoagulant activity Bl6 mice (both wild type and apo EÀ/À), we have demon- such as prothrombin fragment F1 þ 2 and Thrombin-AntiTh- strated that weekly injections of Interleukin-6Fenough to rombin Complex (thrombin generation), and factor VIIc generate a low-level chronic inflammatory respon- (factor VII activation, at least partly) did not. seFcaused a 2–5-fold increase in atherosclerotic lesions.15 We also have some clues about the possible role of Further investigation has suggested that this is due to the hypercoagulability from genetic studies. There have been role of IL-6 in T-cell differentiation, and the generation of clear and compelling associations between venous throm- TH1T helper cells. These cells appear to be critical in the bosis and genetic states such as protein C deficiency (lack of
development of early atherosclerosis in mice. Work is an anticoagulant), factor VLeiden (a procoagulant that is underway to determine if evidence can be gathered to resistant to inactivation), and prothrombin 20210 (increased support a similar role in humans. procoagulant zymogen levels). The presence of genetic A role for inflammation in early atherosclerosis, and variants that result in either increased procoagulation, or results in younger, healthier middle-aged populations have reduced anticoagulation lead to increased risk for venous raised the hypothesis that CRP (and other inflammation thrombosis. However, none of these variants are risk factors markers) are mostly sensitive to the development of lipid- for arterial disease. The exception concerns the atypical rich, mechanically unstable plaques, so-called ‘vulnerable arterial thrombosis occurring in otherwise healthy younger plaques.’ To explore this hypothesis, we collaborated with women. The preponderance of data again indicates that Virmani and co-workers16 at the Armed Forces Institute for there are no compelling data supporting the importance of a Pathology, measuring CRP in post-mortem serum samples pre-existing hypercoagulable state as a major risk factor for from subjects on whom autopsies had been performed and atherothrombotic disease. the type of lesion was well described. The results indicated that CRP was equally elevated in those who died from either thrombosis with vulnerable plaque rupture or fibrous/calci- fied plaque erosion, or stable plaque with no thrombosis. It Association not specific for CVD would appear that inflammation is not particularly asso- One of the emerging issues concerning inflammation and ciated with the development of vulnerable plaques. risk prediction is that these markers are not specific for cardiovascular disease. In fact, in addition to atherothrom- botic disease, markers of inflammation are associated with the presence of, or predictive of the future occurrence of, Hypercoagulable state that predisposes to CVD? virtually all the chronic diseases of older age including heart The pioneering molecular epidemiology of Meade et al,2 at failure, type II diabetes, cancer, frailty, cognitive decline and the Northwick Park Heart Study suggested that a pre-existing dementia, and osteoporosis as well as all-cause mortality. hypercoagulable state might be present in those most at risk These findings raise a number of questions, such as: for myocardial infarction and CVD death. However, ensuing research has provided little support for this as a general is inflammation playing a causal role in all the chronic condition. It does not appear that elevations in factor VIIc diseases of old age?
International Journal of Obesity Inflammation and chronic diseases of old age RP Tracy S33 if ‘yes’, what does this mean with respect to the effect on atherosclerosis of knocking out the fibrinogen independence of these diseases? gene does not support a role for fibrinogen in early is this at least partly the explanation for some of the atherosclerosis. ‘clustering’ of diseases such as is seen with type II diabetes Finally, genetic epidemiological studies can provide some and cardiovascular disease? information in this area as well, since they are not subject to what are the implications for using such markers in the issues of confounding that can complicate the inter- disease prediction and global risk assessment? pretation of phenotypic data. For example, fibrinogen levels are higher in older men and women with greater athero- As the world population ages, answers to these questions sclerotic burden.22 Recent studies from our laboratory as part and others will become even more critical. Increased of the Cardiovascular Health Study have demonstrated that research in these areas is needed. the À455 variant in the fibrinogen gene associated with higher fibrinogen levels (Z Tang, manuscript in preparation) is also associated with greater atherosclerotic burden in older Are markers in the causal pathway? people, supporting data obtained by others. Extending this, Originally, there was a question as to whether markers of we have demonstrated that the À174 C allele of the IL-6 inflammation simply reflected the underlying pathology, or gene, associated with higher IL-6, CRP, and fibrinogen levels, in fact participated in the progression of the atherothrom- is also associated with the lifetime development of athero- botic disease. More recently, as discussed above and sclerosis.23 These data suggest that fibrinogen itself and illustrated in Figure 1, the notion that both of these possible other markers of inflammation are in the causal positions are correct has become generally accepted.5 The pathway, since forms of these genes associated with higher evidence to support the causal pathway model comes from levels of these proteins are more frequent in older people several general areas: laboratory experimentation, animal with greater amounts of atherosclerosis compared to those models, and genetic studies. with little atherosclerosis. For example, a partial list of possible mechanisms directly While there is much work to be done, the bulk of the data linking CRP and fibrinogen to the atherothrombotic process available to date supports the position that at least some of is given in Table 2. These mechanisms have been identified blood markers/mediators of inflammation are in fact in the through lab experimentation, including histological ap- causal pathway of the atherothrombotic process. proaches, and provide considerable indirect support for direct mechanisms for these two factors. In addition, animal model studies can provide some information. For example, as mentioned above, the increase in atherosclerotic burden Role of therapeutics generated by weekly IL-6 injection supports a direct role With the growing body of knowledge about inflammation for IL-6 and/or other proteins whose levels are influenced and vascular disease (and other diseases) comes the desire to by IL-6 in early atherosclerosis. In contrast, the lack of affect the process in a positive way through therapeutics. In fact, the ability to affect atherothrombosis via medications has been supported by a wide variety of studies involving a Table 2 Some potential direct actions of CRP and fibrinogen in athero- number of different therapeutics. Some of the first data in thrombosis and their meaning this area concerned aspirin. In the Physicians Health Study,
CRP actions aspirin therapy was shown to be associated with an Induces monocyte tissue factor expression important reduction in first MI in healthy younger middle- Potent procoagulant aged men. Later analyses, after CRP values were obtained, Colocalizes with complement in plaques showed that the benefit from aspirin increased with higher Activates complement 14 Plaque destabilization CRP levels, independent of lipid level, suggesting an anti- Chemotactic for monocytes inflammatory effect. Foam cell formation Fibrates are a class of drugs that act as PPAR-a agonists, Induces endothelial cell adhesion molecules and in general lower triglycerides via activation of fatty Foam cell formation and platelet deposition 24 Increase LDL uptake by macrophage acid metabolism and raise HDL-cholesterol. Fibrates also Foam cell formation lower fibrinogen by decreasing the ability of proinfla- mmatory cytokines such as IL-1b to activate the inflamma- Fibrinogen actions tory cytokine cascade. The fibrinogen effect has been Increased fibrin formation for a given thrombin amount Resistance to fibrinolysis proposed to be a major part of fibrate-mediated CVD risk Increases plasma viscosity reduction. Clot formation Statins also demonstrate beneficial effects on lipoproteins Increases platelet crosslinking and aggregation (targeted more to LDL-cholesterol decline than triglyceride Clot formation 25 Localizes in atherosclerotic plaque decline), but do not affect fibrinogen. However, statins also lower CRP and have an apparent anti-inflammatory effect.
International Journal of Obesity Inflammation and chronic diseases of old age RP Tracy S34 Further, even in those with relatively low plasma LDL- 11 Ferguson MA, Gutin B, Owens S, Barbeau P, Tracy RP, Litaker M. cholesterol levels, statins demonstrate marked CVD risk Effects of physical training and its cessation on the hemostatic system of obese children. Am J Clin Nutr 1999; 69: 1130–1134. reduction if the CRP level is relatively high,26 again 12 Jarvisalo MJ, Harmoinen A, Hakanen M, Paakkunainen U, Viikari supporting the hypothesis that there may be important J, Hartiala J, Lehtimaki T, Simell O, Raitakari OT. Elevated serum benefits to reducing chronic inflammation. C-reactive protein levels and early arterial changes in healthy These data, and others, have supported the concept of children. Arterioscler Thromb Vasc Biol 2002; 22: 1323–1328. 13 Burke AP, Farb A, Pestaner J, Malcom GT, Zieske A, Kutys R, pleiotropic effects of drugs like aspirin, the statins, and the Smialek J, Virmani R. Traditional risk factors and the incidence of fibrates. Research over the next several years should help us sudden coronary death with and without coronary thrombosis in understand in more detail the potential for anti-inflamma- blacks. Circulation 2002; 105: 419–424. tory therapy in primary and secondary risk reduction. 14 Ridker P, Cushman M, Stampfer M, Tracy R, Hennekens C. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. NEJM 1997; 336: 973–979. 15 Huber SA, Sakkinen P, Conze D, Hardin N, Tracy R. Interleukin-6 Conclusions exacerbates early atherosclerosis in mice. Arterioscler Thromb Vasc Biol 1999; 19: 2364–2367. The last 10 y have seen an enormous surge in research 16 Burke AP, Tracy RP, Kolodgie F, Malcom GT, Zieske A, Kutys R, focused on inflammation and heart disease. In parallel, Pestaner J, Smialek J, Virmani R. Elevated C-reactive protein inflammation (used as a term to represent a broad array of values and atherosclerosis in sudden coronary death: association response systems) has become a topic of interest in a number with different pathologies. Circulation 2002; 105: 2019–2023. 17 Tracy RP, Arnold AM, Ettinger W, Fried L, Meilahn E, Savage P. of different areas of chronic disease including type II The relationship of fibrinogen and factors VII and VIII to incident diabetes, cognitive decline and frailty, among others. These cardiovascular disease and death in the elderly: results from the discoveries are opening up many new opportunities for risk Cardiovascular Health Study. Arterioscler Thromb Vasc Biol 1999; assessment and for therapeutics. These will be increasingly 19: 1776–1783. 18 Ridker P, Hennekens C, Cerskus A, Stampfer M. Plasma concen- important as the population continues to age. tration of cross-linked fibrin degradation product (D-dimer) and the risk of future myocardial infarction among apparently healthy men. Circulation 1994; 90: 2236–2240. 19 Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby References P, Rumley A, Lowe GD. Fibrin D-dimer and coronary heart 1 Virchow R. Die Cellularpathologie in ihrer Begru¨ndung auf physiolo- disease: prospective study and meta-analysis. Circulation 2001; gische und pathologische Gewebelehre 1858. 103: 2323–2327. 2 Meade T, Brozovic M, Chakrabarti R, Haines A, Imeson J, Mellows 20 Cushman M, Lemaitre RN, Kuller LH, Psaty BM, Macy EM, S, Miller G, North W, Stirling Y, Thompson S. Haemostatic Sharrett AR, Tracy RP. Fibrinolytic activation markers predict function and ischaemic heart disease: principal results of the myocardial infarction in the elderly. The Cardiovascular Health Northwick Park Heart Study. Lancet 1986; ii: 533–537. Study. Arterioscler Thromb Vasc Biol 1999; 19: 493–498. 3 Liuzzo G, Biasucci L, Gallimore J, Grillo R, Rebuzzi A, Pepys M, 21 Lowe GD, Rumley A, Sweetnam PM, Yarnell JW, Rumley J. Fibrin Maseri A. The prognostic value of C-reactive protein and serum D-dimer, markers of coagulation activation and the risk of major amyloid A protein in severe unstable angina. N Engl J Med 1994; ischaemic heart disease in the Caerphilly study. Thromb Haemost 331: 417–424. 2001; 86: 822–827. 4 Danesh J, Collins R, Appleby P, Peto R. Association of fibrinogen, 22 Tracy R, Bovill E, Yanez D, Psaty B, Fried L, Heiss G, Lee M, Polak J, C-reactive protein, albumin, or leukocyte count with coronary Savage P, for the CHS Investigators. Fibrinogen and factor VIII, heart disease: meta-analyses of prospective studies. JAMA 1998; but not factor VII, are associated with measures of subclinical 279: 1477–1482. cardiovascular disease in the elderly: results from the Cardiovas- 5 Tracy RP. Inflammation in cardiovascular disease: cart, horse or cular Health Study. Arterioscler Thromb Vasc Biol 1995; 15: 1269– bothFrevisited. Arterioscler Thromb Vasc Biol 2002; 22: 1514–1515. 1279. 6 Macy E, Hayes T, Tracy R. Variability in the measurement of C- 23 Jenny NS, Tracy RP, Ogg MS, Luong le A, Kuller LH, Arnold AM, reactive protein in healthy subjects: implications for reference Sharrett AR, Humphries SE. In the elderly, interleukin-6 plasma interval and epidemiological applications. Clin Chem 1997; 43:52–58. levels and the À174G4C polymorphism are associated with the 7 Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C- development of cardiovascular disease. Arterioscler Thromb Vasc reactive protein and low-density lipoprotein cholesterol levels in Biol 2002; 22: 2066–2071. the prediction of first cardiovascular events. N Engl J Med 2002; 24 Elisaf M. Effects of fibrates on serum metabolic parameters. Curr 347: 1557–1565. Med Res Opin 2002; 18: 269–276. 8 Yudkin JS, Kumari M, Humphries SE, Mohamed-Ali V. Inflamma- 25 Branchi A, Rovellini A, Sommariva D, Gugliandolo AG, Fasoli A. tion, obesity, stress and coronary heart disease: is interleukin-6 Effect of three fibrate derivatives and of two HMG-CoA reductase the link? Atherosclerosis 2000; 148: 209–214. inhibitors on plasma fibrinogen level in patients with primary 9 Tracy RP. Inflammation in cardiovascular disease: cart, horse, or hypercholesterolemia. Thromb Haemost 1993; 70: 241–243. both? Circulation 1998; 97: 2000–2002. 26 Ridker PM, Rifai N, Clearfield M, Downs JR, Weis SE, Miles JS, 10 Ferguson MA, Gutin B, Owens S, Litaker M, Tracy RP, Allison J. Fat Gotto AM, Jr. Measurement of C-reactive protein for the targeting distribution and hemostatic measures in obese children. Am J of statin therapy in the primary prevention of acute coronary Clin Nutr 1998; 67: 1136–1140. events. N Engl J Med 2001; 344: 1959–1965.
International Journal of Obesity