doi: /01.ATV

Transcription

1 Lifestyle and Hemostatic Risk Factors for Ischemic Heart Disease : The Caerphilly Study John W. G. Yarnell, Peter M. Sweetnam, Ann Rumley and Gordon D. O. Lowe Arterioscler Thromb Vasc Biol. 2000;20: doi: /01.ATV Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX Copyright 2000 American Heart Association, Inc. All rights reserved. Print ISSN: Online ISSN: The online version of this article, along with updated information and services, is located on the World Wide Web at: Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Arteriosclerosis, Thrombosis, and Vascular Biology can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular Biology is online at: Downloaded from by guest on February 27, 2013

2 Lifestyle and Hemostatic Risk Factors for Ischemic Heart Disease The Caerphilly Study John W.G. Yarnell, Peter M. Sweetnam, Ann Rumley, Gordon D.O. Lowe Abstract We have recently shown that fibrin D-dimer, tissue plasminogen activator () antigen, von Willebrand factor antigen, fibrinogen, plasma viscosity, and white cell count are associated with subsequent ischemic heart disease (IHD) in men aged 49 to 65 years in the Caerphilly Study from South Wales. We now report the contribution of major lifestyle factors to plasma levels of these new risk predictors for IHD. Results were available for up to 2188 men. The contribution of factors associated with lifestyle (smoking, alcohol, body mass index, leisure and work activity, social class, and use of prescribed medicines) to variation in plasma levels of 8 hemostatic variables was examined. All results were adjusted for other lifestyle variables, age, and time of day. Most hemostatic variables increased with age and smoking habit. Increasing levels of alcohol consumption were associated with increases in and plasminogen activator inhibitor () activity and with decreases in fibrinogen and white cell count.,, fibrinogen (nephelometric), and viscosity were positively associated with body mass index. Increasing levels of leisure activity were inversely associated with D-dimer, von Willebrand factor, nephelometric fibrinogen, and viscosity. Use of prescribed medicines (a marker for chronic illness) was associated with adverse levels of D-dimer, fibrinogen, plasma viscosity, and white cell count.,, and plasma viscosity were associated with blood pressure, cholesterol, and triglycerides but not with lipoprotein(a) or homocysteine. We conclude that several lifestyle factors are associated with hemostatic risk predictors for IHD. Lifestyle modifications may reduce IHD risk partly by altering hemostatic function; large intervention studies are required to test this hypothesis. (Arterioscler Thromb Vasc Biol. 2000;20: ) Key Words: heart disease hemostatic factors lifestyle epidemiological study We have recently shown that several hemostatic variables are associated with incident ischemic heart disease (IHD) in a population sample of men from South Wales (the Caerphilly Study). Measures of fibrinolytic activity included fibrin D-dimer (a marker of fibrin turnover), tissue plasminogen activator antigen (), and plasminogen activator inhibitor activity (). 1 D-dimer appeared an independent predictor of IHD. was also associated with increased risk of IHD, but this association was largely explained by the association of antigen with lipid levels. The association of activity with IHD was nonsignificant. A marker of endothelial dysfunction, von Willebrand factor antigen (), was also independently associated with subsequent risk of IHD. 2 fibrinogen, viscosity, and white cell count were also associated with incident IHD. 3 Fibrinogen has been previously shown to be a risk factor for IHD in numerous prospective studies 4 and has been previously associated with smoking habit, 5 alcohol consumption, 6 and levels of habitual physical activity. 7 We have reported that a heat-precipitation nephelometric assay of fibrinogen was a better predictor of IHD than the Clauss assay of clottable fibrinogen (Sweetnam et al 8 ). Several personal characteristics, both fixed and modifiable, are known or suspected to be associated with cardiovascular disease. 9 These are often broadly grouped under the heading lifestyle factors and include smoking habit, alcohol consumption, body mass index, social class, and habitual physical activity, both at work and during leisure. In this present study, we examine the contribution of these lifestyle factors to the variation in plasma levels of these hemostatic variables in a sample of middle-aged men. We also consider the effects of age, the time of blood sampling, and the use of prescribed medication as an index of chronic illness. Methods Study Populations and Design The original cohort of 2512 men aged 45 to 59 years was recruited between 1979 and 1983, and since then they have been reexamined at 5-year intervals. The men who are the subject of this report are those who were seen at the first reexamination between 1984 and 1988, when they were aged 49 to 65 years. Men of the same age who had moved into the defined geographical area since the original Received March 26, 1999; revision accepted August 7, From the Department of Epidemiology and Public Health (J.W.G.Y.), Queen s University, Belfast, UK; the Medical Research Council Epidemiology Unit (P.M.S.), Cardiff, UK; and the Department of Medicine (A.R., G.D.O.L.), University of Glasgow, Glasgow, UK. Correspondence to John W.G. Yarnell, Department of Epidemiology and Public Health, Queen s University of Belfast, Mulhouse Building, Royal Victoria Hospital Site, Grosvenor Road, Belfast BT12 6BJ, UK. h.porter@qub.ac.uk 2000 American Heart Association, Inc. Arterioscler Thromb Vasc Biol. is available at Downloaded from by guest on February 27, 2013

3 272 Arterioscler Thromb Vasc Biol. January 2000 TABLE 1. Hemostatic Factors: Mean (SD) Values, Medians, and Geometric Means With 95% CIs for Geometric Mean Based on 100, 400, and 900 Men 95% CI* for Geometric Mean Based on Variable Mean SD Median Geometric Mean 100 Men 400 Men 900 Men antigen, antigen, activity, % pool antigen, fibrinogen, fibrinogen, viscosity, White cell count, *Assumes that the SD in the subgroups of 100, 400, and 900 men is the same as in the whole group. recruitment were also eligible to be examined. A total of 2398 men attended the evening clinic, and a fasting blood sample was obtained from 2223 (93%) of them. The general design and methods of the Caerphilly Study have been described elsewhere. 10,11 Briefly, at each examination, the men were invited to attend an afternoon or evening clinic at which a detailed medical and lifestyle history was obtained, the London School of Hygiene and Tropical Medicine chest pain questionnaire 12 was administered, a full 12-lead ECG was recorded, and height, weight, and blood pressure were measured. Alcohol consumption was determined by the use of a self-administered questionnaire checked at the clinic and converted from standard units of volume to milliliters of pure alcohol per week. 13 Habitual physical activity was assessed for work-related activity by use of the Health Insurance Plan questionnaire, 14 and a detailed questionnaire derived from the Minnesota Leisure Time Activity questionnaire 15 was used to estimate the mean daily leisure time energy expenditure. The men were then invited to return, fasting, to an early morning clinic at which a blood sample was taken. Blood Collection, Storage, and Analysis Blood was taken between 7:00 AM and 10:00 AM for 91% of the men, before 7:00 AM for 7%, and between 10:00 AM and 11:00 AM for 2%. Centrifugation of citrated blood was carried out within 1 hour, and samples were stored at 70 C. For D-dimer,,,, and clottable fibrinogen, assays were carried out with the use of these stored samples. One batch of samples was unavailable for the present analysis, so that D-dimer and antigens were measured for a total of 1998 fasting samples. Another batch had been thawed on one occasion and was therefore unsuitable for activity, which was measured for 1569 samples. The measurements were made during 1994, when the plasma had been stored for between 6 and 10 years. ELISAs were used for measurement of D-dimer (AGEN) and (Biopool). activity was measured by use of a chromogenic assay (Chromogenix). fibrinogen was later measured by an automated Clauss assay in a Coag-A-Mate X 2 coagulometer (Organon Teknika) with the use of a separate stored plasma sample. One third of these samples had previously been used for another purpose, so that a total of only 1378 samples were available. Fresh dipotassium edetate anticoagulated samples were used to measure nephelometric fibrinogen, plasma viscosity, and white cell count as described previously. 5 Lipoprotein(a) [Lp(a)] was measured by using a 2-site immunoradiometric assay (Pharmacia). 16 Homocysteine was measured by high-performance liquid chromatography Macroglobulin or 1 -antitrypsin was measured nephelometrically with Beckman antisera kits. Lipid measurements were made by using enzymatic assays as described previously. 11 the lifestyle factors. Therefore, the analyses show how the level of each hemostatic factor varies with any particular lifestyle factor after adjusting for age, time of day, and all the other lifestyle factors. The results are presented, in Tables 2 to 8, as the adjusted means of each hemostatic factor at various levels of the particular lifestyle factor. For the categorical lifestyle factors, such as smoking habit and the use of prescribed medicines, clearly defined categories of adequate size were used. For the continuous variables, such as body mass index and alcohol consumption, the levels were obtained by dividing the variable into 5 equally sized groups. The one exception was age, which was divided into 3 groups: 55, 55 to 59, and 60 years. The distributions of all 8 hemostatic factors have a positive skew, and all were transformed to logarithms. For each, this transformation produced a more nearly symmetric gaussian distribution and stabilized variances. The means in Tables 2 to 8 are thus geometric means. For the categorical variables, tests of significance compared each adjusted mean value with that for a baseline group, such as the men who had never smoked or were taking no prescribed medicine. For the other variables, a test for trend was obtained by replacing the grouped variable with the original continuous value. These were tests for linear trend, with the exception of body mass index, for which a quadratic trend was also fitted. No measure of the variation in the geometric means is given in Tables 2 to 8 in an attempt to retain clarity. Instead, Table 1, which presents the basic characteristics of each hemostatic factor, also gives the 95% CI for geometric means based on 100, 400, and 900 men. These numbers cover the range of sizes of the subgroups, most of which contain 400 men. These CIs assume that the variance in the subgroup is the same as in the whole cohort. This assumption generally holds well after the variance-stabilizing logarithmic transformation. The descriptive statistics shown in Table 1 are given for the maximum number of results available from fasting blood samples for each of the hemostatic tests. The regression analyses are based on 2166 men who had no missing values for any of the nine lifestyle factors and had given a fasting blood sample. Measurements of hemostatic factors were not available for all 2166 men; the numbers available were 2131 for white cell count, 1946 for D-dimer, 1527 for, and 1346 for clottable fibrinogen. The numbers for plasma viscosity and nephelometric fibrinogen were similar to those for white cell count, and the numbers for and were similar to those for D-dimer. Because of the large number of subjects, even very weak associations between the hemostatic and lifestyle factors are statistically significant at the usual 5% level. We have also carried out many tests of statistical significance. To minimize these effects and to try to make the pattern of associations clearer, we have chosen the 1% level as our level of statistical significance. Where a probability is 1%, the exact probability is given, unless P If the probability is 1%, it is deemed to be not statistically significant. Statistical Methods The 8 hemostatic factors were used, in turn, as the dependent Results variables in a series of multiple regression analyses. In each, the Table 1 shows the descriptive data for the 8 hemostatic independent variables were age, Downloaded time of blood from sample, and each of variables. by guest on February 27, 2013

4 Yarnell et al Lifestyle, Hemostasis, and IHD Risk 273 Personal and lifestyle variables showing statistically significant associations with each hemostatic variable after adjustment for all 8 other personal and lifestyle variables. A total of 9 variables were examined: age, smoking habit, alcohol consumption, body mass index (BMI), leisure and work activity, social class, use of prescribed medicines, and time of blood sampling. The Figure summarizes the data shown in Tables 2 to 8 but relates to individual hemostatic factors rather than to individual lifestyle variables. All lifestyle variables shown in the Figure were significantly related to each individual hemostatic factor after adjustment for age, time of blood sampling, and all other lifestyle factors. Table 2 shows that all variables considered (with the exception of white cell count) were associated with age. The majority increased with age, but activity decreased. Venous blood samples were taken after an overnight fast. In order not to disrupt the normal lifestyle of working men, such samples were taken between 4:00 AM and 11:00 AM, giving rise Downloaded from to possible variation in values due to differing circadian rhythms. Table 3 shows that only antigen and plasma viscosity were significantly associated with time of blood sampling; early samples were associated with lower values. Table 4 shows the variation in hemostatic factors with smoking habit. The majority of variables were strongly associated with smoking habit; the strongest associations were for white cell count, fibrinogen, viscosity, and. There were weaker associations with D-dimer and. Dose-dependent relations between the amount of tobacco smoked were usually weaker or absent. Larger effects were seenby between guest on lifetime February abstainers 27, 2013 and exsmokers.

5 274 Arterioscler Thromb Vasc Biol. January 2000 Figure 1. Continued Table 5 shows that associations between the hemostatic factors and alcohol consumption are variable both in magnitude and direction. The strongest associations were the positive ones with and. Negative associations were seen with both clottable fibrinogen and white cell count and, to a lesser extent, with nephelometric fibrinogen. Body mass index is examined in Table 6. Strong positive linear associations were shown for plasma viscosity, nephelometric (but not clottable) fibrinogen, and. showed both strong linear and quadratic associations; the quadratic components reflected the tendency for the rate of increase of to decrease with increasing body mass index. showed only a nonlinear association. Only 2 factors, D-dimer and, showed a statistically significant association with leisure activity as assessed by the Minnesota Leisure Time Activity questionnaire (Table 7). Both decreased as leisure time activity increased. Fibrinogen and plasma viscosity also both decreased as activity increased, but the changes were small and not statistically significant. Work-related physical activity showed no independent relation to any of the hemostatic factors studied (data not shown). Table 8 shows the mean values for the hemostatic factors in subjects who were taking prescribed medicines and in those who were not. Such a measure may be used as a proxy measure for any subjects with long-standing (or, less frequently, short-term) illness, which may affect the variables. This shows that the majority of variables, with the exception of, activity, and, were affected by use of TABLE 2. Variation in Hemostatic Factors With Age (Geometric Mean Values) Age, y Test for trend in adjusted mean values t P NS The geometric mean values are given adjusted for 8 other personal characteristics (time of blood sampling, smoking habit, alcohol consumption, body mass index, leisure and work activity, social class, and use of prescribed medicines. s given are for white cell count. s are similar for nephelometric fibrinogen and plasma viscosity. They are 9% Downloaded lower for D-dimer, from, and antigen, 28% lower for by, guest and on 37% February lower 27, for clottable 2013 fibrinogen.

6 Yarnell et al Lifestyle, Hemostasis, and IHD Risk 275 TABLE 3. Variation in Hemostatic Factors With Time of Day (Geometric Mean Values) Time of Day Before 7:00 AM :00 7:59 AM :00 8:59 AM :00 9:59 AM After 10:00 AM Test for trend in adjusted mean values t P NS NS NS NS NS NS NS indicates not significant. See Table 2 legend. prescribed medication or by the underlying pathology or by both. The relation between social class and hemostatic variables was also examined. Only fibrinogen, measured nephelometrically, and plasma viscosity showed a significant trend with social class (data not shown). To examine the relation between hemostatic factors and other cardiovascular risk markers, we calculated bivariate Pearson correlation coefficients for these variables. Table 9 shows these results. Both and activity were strongly associated with all major cardiovascular risk factors, whereas D-dimer and showed almost no such associations. The acute phase reactants, 2 -macroglobulin and 1 -antitrypsin, showed associations with all hemostatic factors except ; correlated only with 2 -macroglobulin. Lp(a) correlated only with clottable fibrinogen, whereas homocysteine showed only weak correlations with D-dimer,,, and white cell count. The values of the correlation coefficients are given after logarithmic transformation but are very similar for untransformed data. Discussion These results show a series of statistically and clinically significant associations between several modifiable lifestyle variables and certain hemostatic variables, which are both risk predictors for IHD and biologically relevant to its pathogenesis. The advantages of our data set are as follows: (1) it comes from a well-defined cohort study, comprising the majority of middle-aged men in a defined geographical area, Caerphilly 10,11 ; (2) the hemostatic variables studied have all been related to risk of incident IHD in this cohort 1 3 ; and (3) the relations between individual lifestyle variables and hemostatic variables have been clarified by statistical adjustment for other lifestyle variables and also for age, time of blood sample, and prescribed medications. The 95% CIs shown in Table 1 provide a guide for the relative strengths of the trends shown in Tables 2 to 8. The increases with age (between 49 and 65 years) for most variables shown in Table 1 are consistent with the literature for D-dimer, 18,19 antigen, 18,20 antigen, 18,21 fibrinogen, 6 and plasma viscosity. 22 The lack of increase in white cell count 22 and small decrease in activity 18,20 are also consistent with previous data. Diurnal variations (Table 3) in antigen have also been reported. 20 These effects of age and time of blood sample were adjusted for in analyses of lifestyle variables and hemostatic factors. The increases with smoking habit shown in Table 4 are also consistent with the literature for D-dimer, 19 antigen, 23 activity, 6,23,24 antigen, 21 fibrinogen, 4,5,6,18,22 viscosity, 5,18,22 and white cell count. 5,18,22 In longitudinal analyses, however, D-dimer, 1, 1, 2 fibrinogen, 3,4,10 viscosity, and white cell count 3 were associated with subsequent IHD, independent of smoking habit. It has been previously suggested that the association between smoking habit and subsequent IHD may be mediated, at least in part, by these hemostatic variables, 25 especially fibrinogen. 26,27 The present data suggest that the increases in D-dimer,,, fibrinogen, viscosity, and white cell count are partly reversible in exsmokers and that these changes are not due to TABLE 4. Variation in Hemostatic Factors With Smoking Habit (Geometric Mean Values) Never smoked Exsmoker Cigar/pipe smoker /d * /d * /d * 137.1* See Table 2 legend. *P 0.001, P 0.01, and P Downloaded for comparison from of adjusted mean values with those by guest of the never-smoked on February group. 27, 2013

7 276 Arterioscler Thromb Vasc Biol. January 2000 TABLE 5. Variation in Hemostatic Factors With Alcohol Consumption (Geometric Mean Values) Alcohol, ml Pure Alcohol/wk Test for trend in adjusted mean values t P NS NS NS See Table 2 legend. concurrent changes in other lifestyle variables. If so, then the reductions in thrombotic tendency and plasma viscosity may partly explain the rapid reduction in IHD risk that follows cessation of smoking and cannot be attributed to regression of coronary atherosclerosis. 28 Moderate alcohol consumption decreases the risk of IHD; however, heavy alcohol consumption increases the risk of cardiovascular disease, including stroke. 29 The effects of increasing alcohol consumption on hemostatic variables (Table 5) suggest possible explanations for these epidemiological findings. Alcohol consumption was independently associated with decreasing levels of fibrinogen and white cell count, which are consistently associated with IHD risk. 3,30 On the other hand, heavy alcohol consumption was independently associated with increased levels of antigen, as in 2 previous reports from the US Physicians Study 31 and the Edinburgh Artery Study. 32 These 2 studies have also reported that antigen levels are associated with incident stroke, 33,34 and a recent report has associated increased levels of complexes with incident stroke, especially hemorrhagic stroke. 35 Our finding that activity is also related to alcohol consumption, in a manner similar to antigen (Table 5), is consistent with a previous study, 36 raising the possibility that complexes are related to alcohol consumption. Fibrin D-dimer levels were unrelated to alcohol consumption, suggesting that the increased levels of and in high consumers do not result in increasing turnover of cross-linked fibrin in vivo. However, the association of high-dose exogenous as thrombolytic therapy for acute myocardial infarction with hemorrhagic stroke 37 suggests a possible role for alcohol-elevated endogenous in hemorrhagic stroke. On the other hand, it should be noted that raised antigen levels probably do not correlate with functional activity and may relate more to endothelial cell disturbance (such as found with ). Overweight and obesity, assessed by body mass index, are risk factors for IHD and maturity-onset diabetes mellitus. 38 Table 6 confirms that both antigen and activity are strongly related to body mass index, in accordance with the literature In the present study, obese men (body mass index 30 kg/m 2 ) had an 50% increase in activity compared with men of ideal body mass index ( 25 kg/m 2 ), when adjusted for other lifestyle variables. This was accompanied by an 30% increase in antigen and with no increase in cross-linked fibrin turnover in vivo, as assessed by fibrin D-dimer levels. There is some evidence that weight TABLE 6. Variation in Hemostatic Factors With Body Mass Index (Geometric Mean Values) Body Mass Index, kg/m Tests for trend t Linear component only P NS NS NS NS Additional quadratic component t P NS NS NS NS NS NS See Table 2 legend. Downloaded from by guest on February 27, 2013

8 Yarnell et al Lifestyle, Hemostasis, and IHD Risk 277 TABLE 7. Variation in Hemostatic Factors With Leisure Activity (Geometric Mean Values) Leisure kcal/d Test for trend in adjusted mean values t P NS NS NS NS NS NS See Table 2 legend. reduction reduces and levels, 41 suggesting a causal relation. Increasing body mass index was also associated with increasing levels of fibrinogen and plasma viscosity, consistent with the literature, 4,6,16,39,40,42 although, overall, there is little evidence that weight reduction reduces fibrinogen or viscosity levels. 42 Interestingly, body mass index was associated with total (nephelometrically assayed) fibrinogen rather than clottable fibrinogen (Table 6), which may be relevant to our finding that the former assay was a better predictor of incident IHD than the latter in this cohort. 8 Many cohort studies indicate that physical activity, particularly during leisure time, is a protective lifestyle activity against IHD, even in moderate doses. 43 Previous reports from the present cohort 7 and from others 43,44 have associated leisure activity with decreasing fibrinogen levels. In the present study, adjustment for other lifestyle variables reduced this association to nonsignificance at the 1% level. Likewise, the association between fibrinogen and physical activity was no longer statistically significant after adjustment for smoking and social class in the Scottish Heart Health Study. 44 In combination with the conflicting data from intervention studies, 42 a causal association between leisure activity and fibrinogen remains to be proven by further studies. We observed no significant independent association of leisure activity with or levels, which again is consistent with data from intervention studies. 45 In contrast, we observed independent associations of leisure activity with decreased levels of fibrin D-dimer and (Table 7). These novel findings suggest the possibility that leisure activity may reduce IHD risk by decreasing cross-linked fibrin turnover and endothelial disturbance in vivo. However, this hypothesis requires testing in randomized intervention trials of increased leisure activity. A further important personal characteristic of middle-aged men is the prevalence of prescribed medication. Such medication may be prescribed for both acute and chronic illnesses, including symptoms of cardiovascular disease. Table 8 shows that fibrinogen, viscosity, white cell count, and D-dimer were all significantly higher in the men receiving prescribed medication. These findings may reflect effects of the underlying illness, effects of medication, or both. The effects of the underlying illness are probably most important, because all these hemostatic variables are recognized acute-phase reactants, which increase nonspecifically in both vascular and nonvascular illnesses, which are common in comprehensive samples of the general population such as the present cohort. The associations of these variables with the acute-phase reactant proteins, 2 -macroglobulin and 1 -antitrypsin (Table 9), are consistent with this suggestion. Adjustment for prescribed medication (received by almost 50% of men in this cohort, Table 8) is therefore important when considering the effects of lifestyle variables. Table 9 shows that,, nephelometric fibrinogen, and viscosity are correlated with blood pressure and lipids. These findings are consistent with the literature 4,6,22,39 41 and may partly account for the associations of these hemostatic variables with correlated lifestyle variables such as alcohol, obesity, and leisure activity. However, a causal role for blood pressure and lipids in elevation of hemostatic variables remains to be established by large randomized controlled trials of the effects of blood pressure and lipid reduction on hemostatic variables. TABLE 8. Variation in Hemostatic Factors With Prescribed Medicine (Geometric Mean Values) Prescribed Medicine No Yes Test to compare the adjusted mean values t P NS NS NS See Table 2 legend. Downloaded from by guest on February 27, 2013

9 278 Arterioscler Thromb Vasc Biol. January 2000 TABLE 9. Correlation Coefficients Between Hemostatic Factors and Cardiovascular Risk Markers Antigen Antigen Activity Antigen Fibrinogen Fibrinogen Viscosity Count Diastolic BP * Systolic BP 0.07* * * Total cholesterol HDL cholesterol Triglycerides Lp(a) * 0.02 Homocysteine * 0.06* Macroglobulin Antitrypsin BP indicates blood pressure. *P 0.01, P 0.001, and P Lp(a) may be a risk factor for IHD, and it has been suggested that this may be mediated by effects on hemostatic variables: in particular, decreased lysis of intravascular fibrin may result from elevated Lp(a), which competes with plasminogen for binding to fibrin. 46 However, Table 9 shows no significant associations of Lp(a) with fibrinolytic variables; in particular, there was no association with fibrin D-dimer, a marker of turnover of cross-linked fibrin in vivo. The only association of Lp(a) was with clottable (but not nephelometric) fibrinogen, but the degree of association is small. homocysteine may also be a risk factor for IHD in the general population, possibly, in part, because of endothelial disturbance. 47 In the present study, homocysteine correlated weakly with, D-dimer, and white cell count (Table 9). Although consistent with an effect of raised homocysteine on endothelial disturbance, fibrin turnover in vitro, and an inflammatory response, these associations were very weak. In conclusion, our findings are consistent with the hypothesis that lifestyle risk factors for IHD and stroke (smoking, alcohol consumption, obesity, and leisure activity) may operate partly through changes in hemostatic variables that are risk predictors for cardiovascular events. Further large intervention studies are required to examine this hypothesis. In the meantime, these findings suggest that lifestyle modifications may result in decreased thrombotic tendency in the blood and increased ability of blood to flow. This may be relevant to the early decreased risk of cardiovascular events with lifestyle change, which precedes changes in atherosclerosis. Acknowledgments We thank the British Heart Foundation and the Medical Research Council for financial support. References 1. Lowe GDO, Yarnell JWG, Sweetnam PM, Rumley A, Thomas HF, Elwood PC. Fibrin D-dimer, tissue plasminogen activator, plasminogen activator inhibitor, and the risk of major ischemic heart disease in the Caerphilly study. Thromb Haemost. 1998;79: Rumley A, Lowe GDO, Sweetnam PM, Yarnell JWG, Thomas HF, Ford RP. Factor VIII, von Willebrand factor and the risk of major ischemic heart disease in the Caerphilly Study. Br J Haematol. 1999;105: Yarnell JWG, Baker IA, Sweetnam PM, Bainton D, O Brien JR, Whitehead PJ, Elwood PC. viscosity and white cell count are major risk factors for ischemic heart disease. Circulation. 1991;83: Ernst E, Resch KL. Fibrinogen as a cardiovascular risk factor: a metaanalysis and review of the literature. Ann Intern Med. 1993;118: Downloaded from 5. Yarnell JWG, Sweetnam PM, Rogers S, Elwood PC, Bainton D, Baker IA, Eastham R, O Brien JR, Etherington NW. Some long term effects of smoking on the hemostatic system: a report from the Caerphilly and Speedwell Collaborative Surveys. J Clin Pathol. 1987;40: Meade TW, Chakrabarti R, Haines AP, North WRS, Stirling Y. Characteristics affecting fibrinolytic activity and plasma fibrinogen concentrations. BMJ. 1979;1: Elwood PC, Yarnell JWG, Pickering J, Fehily AM, O Brien JR. Exercise, fibrinogen, and other risk factors for ischemic heart disease: Caerphilly Prospective Heart Disease Study. Br Heart J. 1993;69: Sweetnam PM, Yarnell JWG, Lowe GDO, Baker IA, O Brien JR, Rumley A, Etherington MD, Whitehead PJ, Elwood PC. The relative power of heat-precipitation nephelometric and clottable (Clauss) fibrinogen in the prediction of ischemic heart disease: the Caerphilly and Speedwell studies. Br J Haematol. 1998;100: Poulter N, Sever P, Thom S, eds. Cardiovascular Disease: Risk Factors and Intervention. Oxford, UK: Radcliffe; Sweetnam PM, Thomas HF, Yarnell JWG, Beswick AD, Baker IA, Elwood PC. viscosity and the 10-year incidence of ischemic heart disease. Eur Heart J. 1996;17: Bainton D, Miller NE, Bolton CH, Yarnell JWG, Sweetnam PM, Baker IA, Lewis B, Elwood PC. triglyceride and high density lipoprotein cholesterol as predictors of ischemic heart disease in British men: the Caerphilly and Speedwell collaborative heart disease studies. Br Heart J. 1992;68: Rose GA, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular Survey Methods. 2nd ed. Geneva, Switzerland: World Health Organization; Yarnell JWG, Fehily AM, Milbank JE, Sweetnam PM, Walker CL. A short dietary questionnaire for use in an epidemiological survey: comparison with weighed dietary records. Hum Nutr Appl Nutr. 1983;37A: Shapiro S, Weinblatt E, Frank CW, Sager RV. The H.I.P. study of incidence and prognosis of coronary heart disease: preliminary findings on incidence of myocardial infarction and angina. J Chronic Dis. 1965; 18: Taylor HL, Jacobs DR, Schucker B, Knudsen J, Leon AS, de Backer G. A questionnaire for the assessment of leisure time physical activities. J Chronic Dis. 1978;31: Durrington PN, Ishola M, Hunt LP, Arrol S, Bhatnagar D. Apolipoproteins (a), A-I, and B, and parental history in men with early onset ischemic heart disease. Lancet. 1988;1: Ubbink JB, Vermaak WJH, Bissbort S. Rapid high-performance liquid chromatographic assay for total homocysteine levels in human serum. J Chromatogr. 1991;565: Rumley A, Lowe GDO, Lee AJ, Tunstall-Pedoe HD. Effect of age, sex and menopause on fibrinolytic variables. Fibrinolysis. 1992;6(suppl 3):73. Abstract. 19. Lee AJ, Fowkes GR, Lowe GDO, Rumley A. Determinants of fibrin D-dimer in the Edinburgh Artery Study. Arterioscler Thromb Vasc Biol. 1995;15: Eliasson M, Evrin P-E, Lundblad D, Asplund K, Ranby M. Influence of gender, age and sampling time on plasma fibrinolytic variables and fibrinogen. Fibrinolysis. 1993;7: Conlan NG, Folsom AR, Finch A, Davis CE, Sorlie P, Marcucci G, Wu KK. by guest Associations on February of factor 27, VIII2013 and von Willebrand factor with age, race,

10 Yarnell et al Lifestyle, Hemostasis, and IHD Risk 279 sex, and risk factors for atherosclerosis: the Atherosclerosis Risk in Communities (ARIC) Study. Thromb Haemost. 1998;79: Lowe GDO, Smith WCS, Tunstall-Pedoe HD, Crombie IK, Lennie SE, Anderson J, Barbenel JC. Cardiovascular risk and haemorheology: results from the Scottish Heart Health Study and the MONICA Project, Glasgow. Clin Hemorheol. 1988;8: Eliasson M, Asplund K, Evrin P-E, Lundblad D. Relationship of cigarette smoking and snuff dipping to plasma fibrinogen, fibrinolytic variables and serum insulin: the Northern Sweden MONICA Study. Atherosclerosis. 1995;113: Allen RA, Kluft C, Bronmer EJP. Effect of chronic smoking on fibrinolysis. Arteriosclerosis. 1985;85: Yarnell JWG. The role of hemostatic factors in cardiovascular disease: a review. Primary Cardiol. 1992;18: Meade TW, Imeson J, Stirling Y. Effect of changes in smoking and other characteristics on clotting factors and the risk of ischemic heart disease. Lancet. 1987;2: Kannel WB, D Agostino RB, Belanger AJ. cigarette smoking, and risk of cardiovascular disease: insights from the Framingham study. Am Heart J. 1987;113: Royal College of Physicians. Smoking or Health? London, UK: Pitman; Beaglehole R, Jackson R. Alcohol, cardiovascular diseases and all causes of death: a review of the epidemiological evidence. Drug Alcohol Rev. 1992;11: Danesh J, Collins R, Appleby P, Peto R. Association of fibrinogen, C-reactive protein, albumin or leukocyte count with coronary heart disease. JAMA. 1998;279: Ridker PM, Vaughan DE, Stampfer MJ, Glynn RJ, Hennekens CH. Association of moderate alcohol consumption and plasma concentration of endogenous tissue type plasminogen activator. JAMA. 1994;272: Lee AJ, Flanagan PA, Rumley A, Fowkes FGR, Lowe GDO. Relationship between alcohol intake and tissue-plasminogen activator antigen and other hemostatic factors in the general population. Fibrinolysis. 1995;9: Ridker PM, Hennekens CH, Stampfer MJ, Manson JE, Vaughan DE. Prospective study of endogenous tissue plasminogen activator and risk of stroke. Lancet. 1994;343: Smith FB, Lee AJ, Fowkes FGR, Price JF, Rumley A, Lowe GDO. Hemostatic factors as predictors of ischemic heart disease and stroke in the Edinburgh Artery Study. Arterioscler Thromb Vasc Biol. 1997;17: Nilsson TK, Jansson JH, Boman K. t-pa complex as a risk factor for the development of a first stroke. Fibrinolysis. 1998;12(suppl 1):21. Abstract. 36. Hendriks HF, Veenstra J, Velthuis-te Wierik EJ, Schaafsma G, Kluft C. Effect of moderate dose of alcohol with evening meal on fibrinolytic factors. BMJ. 1994;308: Collins R, MacMahon S, Flather M, Baigent C, Remvig L, Mortensen S, Appleby P, Godwin J, Yusuf S, Peto R. Clinical effects of anticoagulant therapy in suspected acute myocardial infarction: systematic overview of randomised trials. BMJ. 1996;313: Jousilhti P, Tuomilehto J, Vartianen E, Pekkanen J, Puska P. Body weight, cardiovascular risk factors and coronary mortality. Circulation. 1996;93: Sundell IB, Nilsson TK, Ranby M, Hallmans G, Hellsten G. Fibrinolytic variables are related to age, sex, blood pressure, and body build measurements: a cross-sectional study in Nonsjö, Sweden. J Clin Epidemiol. 1989;42: Eliasson M, Evrin P-E, Lundblad D. Fibrinogen and fibrinolytic variables in relation to anthropometry, lipids and blood pressure: the Northern Sweden MONICA Study. J Clin Epidemiol. 1994;46: Juhan-Vague I, Alessi MC. Fibrinolysis and risk of coronary artery disease. Fibrinolysis. 1996;10: Ernst E, Resch KL. Therapeutic interventions to lower plasma fibrinogen concentration. Eur Heart J. 1995;16: Connelly JB, Cooper JA, Meade TW. Strenuous exercise, plasma fibrinogen and factor VII activity. Br Heart J. 1992;67: Lee AJ, Smith WCS, Lowe GDO, Tunstall-Pedoe H. fibrinogen and coronary risk factors: the Scottish Heart Health Study. J Clin Epidemiol. 1990;43: Lowe GDO, Small M. Stimulation of endogenous fibrinolysis. In: Kluft C, ed. Tissue-Type Plasminogen Activator t-pa: Physiological and Clinical Aspects, Volume II. Boca Raton, Fla: CRC Press; 1988: Scanu AM, Scandiani L. Lipoprotein (a): structure, biology and clinical relevance. Adv Intern Med. 1991;36: Boushey CJ, Beresford SAA, Omenn GS, Motulsky AG. A quantitative assessment of plasma homocysteine as a risk factor for vascular disease. JAMA. 1995;274: Downloaded from by guest on February 27, 2013