Protein Z levels and vascular thrombotic diseases: A meta-analysis
F Sofi, F Cesari, S Fedi1, GJ Broze2, R Abbate, GF Gensini
1University of Florence, Florence, Italy, 2Washington University, Division of Hematology, St. Louis, United States of America
Topic: Risk factors and risk prediction
Objective: The relationship between protein Z levels and thrombosis is controversial. We performed a systematic review and meta-analysis of all the available studies to assess the association between protein Z and vascular thrombotic diseases.
Methods: We conducted an electronic literature search through MedLine, Embase, Google Scholar, Web of Science, The Cochrane Library, bibliographies of retrieved articles and abstracts of congresses up to May, 2009. Studies were included if they analysed protein Z levels in patients with vascular thrombotic diseases.
Results: After the review process 28 case-control studies (33 cohort of patients) were included in the final analysis. These studies included 4,218 patients with thrombotic diseases as compared to 4,778 controls. The cumulative analysis for all the studies under a fixed-effects model showed that patients with low protein Z levels are at increased risk of occurring a thrombotic event (OR 1.87, 95%CI 1.62–2.15; p>0.00001). By subgrouping studies according to the different type of the disease, low protein Z levels were found to be a significant risk factor for arterial vascular diseases (OR 1.86, 95%CI 1.51–2.27; p>0.00001), and pregnancy complications (OR 3.42, 95% CI 2.51–4.66; p>0.00001) whereas only a limited significance for venous thromboembolic diseases (OR 1.28, 95%CI 1.00–1.65; p=0.05) was reported.
Conclusion: This is the first meta-analysis that attempted to evaluate the role of protein Z on the occurrence of thrombotic events. The present results suggest, in an overall population of more than 4,000 patients, a possible implication for low levels of this coagulatory protein on the pathogenesis of thrombosis.
P53
Differences in the development and outcomes of acute coronary syndrome in men and women (according to the RECORD registry data)
I I Shevchenko1, N A Gratsiansky2, A D Erlikh2
1Municipal Emergency Care Hospital, # 10 Voronezh, Russian Federation,2Research Institute for Physicochemical Medicine, Moscow, Russian Federation
Topic: Risk factors and risk prediction
Purpose: To study statistical differences in the development and outcomes of acute coronary syndrome (ACS) between groups of men and women at emergency coronary care units.
Population and methods: the study group researched data of 796 consecutive patients (pts) with ACS (341 women and 455 men), who were admitted to emergency coronary care units (the RECORD registry) at 14 cardiac centres.
Results: The following ACS complications occurred in the group in general: Killip I 641 pts (80,5%), including 389 men (85,5%) and 252 women (73,9%) (?=0,000); Killip II 100 pts (12,56%), including 44 men (9,7%) and 56 women (16,4%) (?=0,0049); Killip III 32 pts (4,02%), including 10 men (2,19%) and 22 women (6,45%) (?=0,0029); Killip IV 20pts (2,51%), including 10 men (2,19%) and 10 women (2,93%) (?=0,41%). Acute myocardial infarction (AMI) developed after hospital admission in 34 pts (4,27%), including 11 men (2,42%) and 23 women (6,74%) (?=0,003). In-hospital death occurred in 56 cases (7,04%), which included 20 men (4,4%) and 36 women (10,55%) (?=0,0008). Early mortality was the highest: 27 pts (10 men and 17 women) died in the first 24 hours (48,21% of the total mortality). Kaplan-Meier survival curves are significantly different for men and women in the first 24 hours (Gehan-Wilcoxon test 0,007). In general, however, in-hospital mortality curves are not significantly different for men and women (Fig.).
Conclusions: 1. Women significantly more often developed complications in the acute stage according to Killip classification, as well as in-hospital AMI. 2. Mortality in women was higher and significantly different from that in men in the first 24 hours; in general, however, in-hospital mortality curves are not significantly different for men and women.
Cumulative Proportion Surviving
P54
The influence of metabolic syndrome in patients with stable coronary heart disease in achieving the dual target of low LDL cholesterol and low c-reactive protein
C Varounis1, LS Rallidis1, V Souridis1, A Charalampopoulos1, M Zolindaki2, C Kottakos1, K Tsitlakidis1, M Anastasiou-Nana1
1Attikon University Hospital, Athens, Greece, 2Biochemistry Laboratory, General Hospital of Nikea, Piraeus, Greece
Topic: Risk factors and risk prediction
Purpose: Metabolic syndrome (MS) is associated with higher risk for recurrence of coronary events in patients with coronary artery disease (CAD). On the contrary, it has been shown (PROVE IT TIMI-22 trial) that achieving the dual target, i.e. a low density lipoprotein cholesterol (LDL-chol) >100 mg/dl and a C-reactive protein (CRP) >2 mg/l provides the best long term survival in coronary patients. The aim of this study was to investigate the proportion of patients with MS and CAD who achieve the dual goal.
Methods: We recruited 576 consecutive patients =75 years (mean age=62±23 years, 85% men) with stable CAD from outpatient cardiologic clinic. The presence of risk factors was recorded and fasting lipids and hs-CRP were determined.
Results: One hundred fifty-three patients (26.6%) achieved the dual target. The vast majority of patients (97%) were on hypolipidaemic treatment (95% on statins). Age, gender, hypertension, diabetes mellitus and history of myocardial infarction had no impact on reaching the dual target. However, current smokers had 41% less odds in reaching the double target compared to non-smokers (Odds ratio [0R]: 0.59, 95% confidence intervals [CI]: 0.38-0.91, p=0.02). Patients having MS had 51% less odds in reaching the dual target (OR: 0.49, 95% CI: 0.32–0.76, p=0.001). Only 19% of patients with MS reached the dual target and MS remained a strong and independent predictor of not reaching the dual target after adjustment for age, gender, smoking status and treatment with statins (OR: 1.78, 95% CI: 1.10–2.89, p=0.02).
Conclusions: Patients with stable CAD and MS have a lower probability in reaching the dual target of low LDL-chol and low CRP levels and this may be due to the larger inflammatory burden of MS.
P55
Twenty-year outcome prediction in cardiac patients: relative importance of clinical features and measures of left ventricular systolic and diastolic function
I P Clements, CG Scott, DO Hodge
Mayo Clinic Rochester, United States of America
Topic: Risk factors and risk prediction
Purpose: To investigate if measures of left ventricular (LV) systolic and diastolic function add to clinical history and vital signs as determinants of long-term mortality in cardiac patients.
Patients and methods: Between 1987 and 1998, 778 patients aged =45 underwent resting equilibrium radionuclide angiocardiography (ERNA) for cardiac evaluation. The following variables were recorded at the time of ERNA: age, resting heart rate, blood pressure, gender, medication usage, the presence or absence of a history of smoking, hypertension, family history of coronary disease, diabetes mellitus, myocardial infarction, hyperlipidemia, valvular heart disease, permanent pacemaker insertion and coronary artery revascularization. From the ERNA LV ejection fraction, LV end-systolic and end-diastolic volume index (ml/m2) and peak early rapid filling rate (ml/m2/s) were determined. Date of death was determined from social security records.
Results: Over the 20-year study period 457 deaths occurred. Multivariate analysis identified the presence of the following variables to be associated independently with increased mortality: age (6% absolute increase per year increment in age), male sex (18% increase), diabetes mellitus (90% increase), smoking (31% increase), nitrate usage (24% increase), digoxin usage (25% increase), calcium-channel blocker usage (31% increase), heart rate (19% increase with each 10 beat/min increment), LV end-systolic volume (6% increase with each 10ml/m2 increment). When LVEF was = 0.50, age, male sex, diabetes mellitus, smoking, nitrate usage and heart rate remained independent variables associated with increased mortality. When LVEF was <0.50, LV EF became a significant independent predictor of mortality (25% increase with 0.10 LV EF decrement) and age, male sex, diabetes mellitus, hypertension, and increased LV end-systolic volume remained independently associated with increased mortality.
Conclusions: Over a twenty-year period in cardiac patients: 1. Clinical history provided the major predictors of mortality. 2. Remarkably the resting heart rate had independent predictive value for mortality. 3. When LV EF was preserved measures of LV systolic and diastolic function did not independently predict mortality. 4. When LV EF was reduced, LV systolic function but not diastolic function independently predicted mortality.
P56
Prognostic value of traditional cardiovascular risk factors and vascular disease markers after an acute coronary syndrome in patients receiving optimal medical treatment
M Lafitte, YP Pucheu, M Riedel, T Couffinhal
University Hospital of Bordeaux - Hospital Haut Leveque, Department of Cardiology, Pessac, France
Topic: Risk factors and risk prediction
Background: After an acute coronary syndrome (ACS), intensive pharmacologictherapy and lifestyle intervention (optimal medical therapy, OMT) has proved its efficacy in reducing cardiovascular (CV) events. However, even in populations who reach recommended secondary prevention goals, a subset of patients still experiences CV events.
Aim: To identify biological or cardiovascular markers predicting the residual risk of CV events in post-ACS patients receiving OMT.
Patients and methods: 957 patients with an ACS benefited from an intensified intervention focusing on evaluating risk factors and atherosclerosis lesions, and on optimizing treatment and education. OMT and a therapeutic and dietary education program were started during hospitalization for ACS and adjusted at 3 months. Traditional risk factors and cardiovascular markers (intima-media thickness measurement (IMT), carotidatheroma, Ankle Brachial Index (ABI) and the number of coronary arteries with a>50% stenosis) were evaluated 3 months after the ACS.
Results: At a median follow-up of 20 months, more than 80% of the patients reached recommended secondary preventiongoals. 99 total CV events were recorded during follow-up (CV death, ACS, stroke, heart failure, revascularization). In this population, diabetes was the only CV risk factor significantly associated with CV events in multivariate analysis (including age, sex, hypertension, LDL cholesterol, smoking, metabolic syndrome, hsCRP); HR 1,61(1,09–2,39), p>0,017. In univariate analysis of CV disease markers, peripheral arterial disease (PAD defined as ABI>0,9), carotid plaque> 50%, or 3-coronary vessel disease were significantly associated with CV events, whereas IMT <0,7mm, myocardial ischemia (SPECT) or ejection fraction >40% were not. In multivariate analyses including risk factors and CV disease markers, only PAD remained significantly associated with CV events; HR 1,83 (1,02–3,31), p=0,04. The number of vascular beds involved was associated with poorer prognosis: HR for 3 vascular-beds disease 3,85 (1,72–8,63) p= 0,001, given 1 vascular-beddisease as a reference group.
Conclusion: In post-ACS patients with optimized control of risk factors, PAD and atheroma burden represent powerful prognostic markers of CV events, while diabetes remains the only independent marker of CV events among traditional risk factors.
P57
The lower BMI the higher mortality among 75-year-olds within the upper half of the fasting plasma glucose spectrum
G Nilsson1, J Ohrvik2, P Hedberg3
1Uppsala University, Centre for Clinical Research, Uppsala, Sweden, 2Karolinska Institutet, Department of Medicine, Stockholm, Sweden, 3Central Hospital, Department of Physiology, Vasteras, Sweden
Topic: Risk factors and risk prediction
Background: An association between low BMI and all-cause mortality (the “obesity paradox”) is found among patients with known atherosclerotic diseases. Hyperglycemia predisposes to atherosclerotic diseases. Therefore, we hypothesized that among subjects with high fasting plasma glucose (FPG) there is an association between low BMI and all-cause mortality.
Study population: A random sample of 75-year-olds in a general population were offered a health survey; 70% (210 men and 222 women), including 34 cases with known diabetes, accepted the invitation. A stratified analysis of all-cause mortality was used for subjects with FPG below and above the median FPG (5.9 mmol/L).
Results: Median (interquartile range) for BMI was 24.9 (23.0–26.9) and 26.1 (24.2–28.4) among subjects with FPG below and above the median FPG (P>0.001 for difference). The 10-year mortality was 32% and 42% for subjects with FPG below and above the median FPG, respectively. Results of Cox analyses, with and without adjustment for sex, known hypertension, prior myocardial infarction and prior stroke are presented in the Table 1. Including the interaction between BMI and dichotomized FPG in a Cox regression analysis demonstrated a significant (P=0.020) difference in the association of BMI to survival between individuals FPG below and above the median.
Conclusion: Within a general population of 75-year-old people, there is an association between BMI and all-cause mortality in subjects above but not below the median FPG.
Unadjusted and adjusted Hazard ratios for all-cause mortality in relation to BMI, stratified for fasting plasma glucose below and above the median.
Plasma glucose
Hazard ratio; HR (95% CI)/unit BMI
P-value
Low (FPG=5.9 mmol/L)
1.02 (0.96–1.09)
0.525
Low adjusted
1.00 (0.93–1.07)
0.953
High (FPG>5.9 mmol/L)
0.92 (0.86–0.98)
0.006
High adjusted
0.91 (0.85–0.97)
0.003
P58
GGT measurement in primary care: a coherent indicator of cardiovascular risk?
CM Conlon1, C Kelleher1, I Dawkins2, C O Loughlin2, M Ledwidge2, K Mc Donald2
1University College Dublin, School of Public Health, Physiotherapy & Population Science, Dublin, Ireland, 2St Vincent's University Hospital, Dublin, Ireland
Topic: Risk factors and risk prediction
The STOP HF Investigators
Background: Gamma-Glutamyltransferase (GGT) is back on the radar, now as a marker of cardiovascular (cv) risk; rather thank simply an indicator of liver disease. Both the Vorarlberg Health Monitoring and Promotion Program and the British Regional Heart Study has linked GGT to mortality and cardiovascular events in large general population and in high risk hospitalised cohorts. Despite the convincing findings of these studies, there is a dearth of data on the associations between GGT and groups with conventional cv risk factors in the Primary Care setting.
Objective: To examine the factors associated with elevated GGT in an Irish Primary Care population selected based on cv risk factors.
Methodology: We explored the baseline data set of the ongoing STOP HF Study, a prospective study of individuals based in primary care with defined cv risk factors and no known ventricular dysfunction. We conducted logistic regression using GGT above and below the 75th percentile GGT as the binary outcome for both gender (males: 49u/L; females 38u/L), to identify multivariate (MV) associates of higher GGT.
Results: Complete data was available in 879 individuals (mean age 65 years, 47%male, 66% hypertension, 49% hypercholesterolemia, 25% obesity, 16% coronary artery disease–49% of whom had prior MI, 15% smoking; 11% diabetes). The most frequently applied medical therapies were Statins (63%), Anti-platelet (49%), Beta-blockers (30%), ACE-Inhibitors (28%), AIIA (26%), Thiazides (20%), Calcium Antagonists (16%) and Anti-diabetic Agents (11%). MV associates of higher GGT [Exponential beta-coefficient (95% confidence interval); p-value] in males were younger age [.97(.96-.99); .009], higher diastolic blood pressure (BP) [1.05 (1.02–1.07); .000] total cholesterol [1.99 (1.19–3.39); .001] and HsCRP [2.01 (1.123.57); .018] and lower urea [.75 (.63-.89); .001] and HDL [.33 (.18 - .61); .000] and in females, higher body mass index [1.08(1.03–1.13); .001] and systolic BP [1.01(1.00–1.02); .049] and more beta blocker use [1.49(1.27–1.87); .016].
Conclusions: Exploration of associates of GGT in this at-risk cohort corroborates recent findings of large general population cohorts. We demonstrate that independently and even within its normal ranges, GGT is associated with markers of cv risk in a primary care population. Particularly in males, GGT appears to be a coherent risk factor associated with incipient underlying disease, in keeping with mechanistic evidence suggesting its role in atherogenesis. GGT measurement is an easily accessible and inexpensive biomarker for cardiovascular risk assessment that may have utility in further algorithms.
P59
Association between cardiac arrhythmias and undiagnosed obstructive sleep apnea (OSA) in the general population: The Akershus Sleep Apnea Project (ASAP)
S Namtvedt1, A Randby1, G Einvik2, V Somers3, T Omland1
1Universitetet i Oslo, Oslo, Norway, 2Akershus University Hospital, Department of Medicine, Lorenskog, Norway, 3Mayo Clinic, Rochester, United States of America
Topic: Risk factors and risk prediction
Purpose: Sleep clinic derived data suggest that the prevalence of cardiac arrhythmias is high in patients with obstructive sleep apnea (OSA) (1848%), and may contribute to the increased cardiovascular morbidity and mortality associated with OSA However, in the general population OSA is often undiagnosed, and the prevalence of arrhythmias in European subjects with undiagnosed OSA is unknown. The aim of this study was therefore to estimate the prevalence of cardiac arrhythmias according to OSA status in the general population.
Methods: 541 subjects (mean age 48.3, 54.8 % male) were invited from a population-based cohort of 30 000 subjects aged 30–65 years in Norway. Subjects classified as being high-risk according to the Berlin questionnaire were over sampled and included in the study. In-hospital polysomnography and Holter recordings were performed. The occurrence of ventricular arrhythmias (> 5/h premature ventricular contractions (PVC), any bigeminy, trigeminy or-, nonsustained ventricular tachycardia (NSVT)), supraventricular arrhythmias (premature atrial contractions (PAC), atrial fibrillation) and conduction delay arrhythmias (sinus pause, AV-block) were recorded. Multivariate logistic regression analyses, adjusted for standard cardiovascular risk factors, were performed to evaluate the association between apnea hypopnea index (AHI) and arrhythmias.
Results: The prevalence of subjects who had tachy- or bradyarrhythmias was generally lower than previously observed in sleep clinic samples. We observed a higher prevalence of premature ventricular contractions and ventricular arrhythmias in subjects with OSA than those without (Table 1). This association was attenuated but still significant (p>0.05) after adjusting for confounding variables.
Conclusion: Undiagnosed OSA in the general population was associated with a higher prevalence of both premature ventricular contractions and ventricular arrhythmias, although the frequency of these arrhythmias was lower than that reported in previous studies.
Subjects with arrhythmia (%)
PVC = 5/h
Ventricular arrhytdmias
Pause>2s
PAC =5/h
Atrial fibrillation
AHI>5
5.2
5.6
6.9
48.7
0
AHI=5
13.3∗
12.9∗
9.8
46.2
1.4
Ventricular arrhythmias are defined as bigeminy, trigeminy, or NSVT. ∗p>0.01.
P60
The ratio of non-fatal to fatal cardiovascular diseases and implications for risk charts
I Van Dis1, D Kromhout2, JM Geleijnse2, JMA Boer1, WMM Verschuren1
1National Institute for Public Health and the Environment, Bilthoven, Netherlands, 2Wageningen University, Wageningen, Netherlands
Topic: Risk factors and risk prediction
Background: In Europe, the SCORE risk charts are used for the prediction of an individual's 10-year cardiovascular risk, based on mortality from cardiovascular diseases (CVD). However, the WHO MONICA project demonstrated that in Western Europe CVD mortality rates declined more strongly than non-fatal CVD event rates in the nineteen eighties. Due to the strong decline in CVD mortality rates observed since the nineteen seventies, risk charts based on CVD mortality only, may not reveal the impact of risk factors on the total burden of CVD.
Purpose: To study the trends in CVD morbidity and mortality and their ratio using recently collected population-based data in the Netherlands.
Methods: During 1993–1997, risk factor data were obtained from 14,500 men and women aged 35–65 years, who were randomly selected from three Dutch towns (Monitoring Project on Risk Factors for Chronic Diseases). Information on non-fatal CVD was obtained from the National Hospital Discharge Register and on fatal CVD from Statistics Netherlands. Non-fatal and fatal CVD was defined as the occurrence of the first non-fatal or fatal CVD event. We composed two clusters of diseases and mortality: one based on all ICD-codes as applied in the SCORE-project (SCORE-CVD) and one on ICD-codes for myocardial infarction and stroke only (ICD-codes I21, I60, I61, I63, I64, I66) (AMI+CVA).
Results: 10-year SCORE-CVD mortality rate declined from 1.2% in participants recruited in 1993 to 0.6% in those recruited in 1997. For AMI+CVA, mortality rates declined from 0.7% to 0.4% respectively. On average, over the whole period (1993–1997) SCORE-CVD mortality rate amounted to 0.7% and AMI+CVA mortality to 0.4%. The rate of the combined endpoints non-fatal and fatal CVD was 9% for SCORE-CVD and 3% for AMI+CVA. The ratio of non-fatal plus fatal CVD to fatal CVD was 13 for SCORE-CVD and 8 for AMI+CVA. These ratios decreased with increasing age from 19 at age 35–45 to 11 at age 55–65 for SCORE-CVD and from 20 to 8 for AMI+CVA.
Conclusions: In the Netherlands CVD mortality rates continued to decline in recent years, which resulted in high ratios for non-fatal plus fatal CVD to fatal CVD. To take the total burden of disease into account, up-to-date risk charts should use non-fatal and fatal CVD combined as an endpoint.
P61
Educational attainment influences cardiovascular disease risk in middle-aged men
R Siren1, J Eriksson2, H Vanhanen3
1Health Centre of City of Helsinki, Helsinki, Finland, 2University of Helsinki, Helsinki, Finland, 3The Sosial Insurance Institution (Kela), Helsinki, Finland
Topic: Risk factors and risk prediction
Purpose: To evaluate the effectiveness of systematic health counseling on cardiovascular disease risk factors in middle-aged men with different levels of education.
Methods: In 2006 all men aged 40 living in Helsinki were invited to a cardiovascular disease (CVD) risk assessment visit at their local health care centre. During the appointment with trained nurses, participants were interviewed about their lifestyle. Blood pressure, height, weight, and waist circumference were measured according to standard technique; BMI was calculated. Blood samples were drawn for fasting lipids and glucose. All participants completed the CVD risk index questionnaires (The Modified Cardiovascular Risk Index of the North Karelia project). Of the 4274 men that were invited, 1454 (34%) participated and provided baseline data. 471 men with a CVD risk index of =4.5 received lifestyle counseling based upon their own risk profile. In 2008 a total of 430 were identified in the local health care centre register and were invited to a follow-up visit. Of these, 200 men participated; the same measurements were made as in 2006. Level of educational attainment was available for 82% of the men.
Results: For the analyzes we divided the participants in 3 groups by the duration of education; >9 years n=26; (16%) group A, 10 to 12 years n=76; (46%) group B, and =13 years n=62; (38%) group C. At baseline in 2006 the means of the CVD risk indexes of the groups A, B, and C were 7.09, 6.30 and 5.83, respectively. The difference of the means for groups A and B was 0.79 (95% CI 0.002 to 1.59; p=0.05), for groups B and C 0.47 (95% CI 0.01 to 0.93; p=0.044), and for groups A and C 1.26 (95% CI 0.50 to 2.03; p=0.002). In 2008 at the follow-up visit the means of the CVD risk indexes of the groups were 6.77,5.76 and 4.80, respectively. The respective difference of the means for groups A and B was 1.01 (95% CI 0.02 to 1.99; p=0.046), for B and C 0.96(95% CI 0.29 to 1.64; p=0.005), and for A and C 1.97 (95% CI 1.11 to 2.83; p>0.001). The differences within the groups by means in the 2006 and 2008 visits were as follows: for group A 0.32 (95% CI –0.59 to 1.24; p=0.469), for group B 0.54 (95% CI 0.11 to 0.97; p=0.014), and for group C 1.03 (95% CI 0.61 to 1.45; p>0.001).
Conclusions: Subjects with the lowest educational attainment tended to have the highest CVD risk index and impact of the health counseling was less effective among them compared with the higher educated groups. The methods used for health counseling should be person oriented taking into account the socioeconomic background of the subjects.
P62
Total/high density lipoprotein cholesterol ratio improves prediction of cardiovascular death by SCORE in men, but not in women - improves prediction of ischemic heart disease, but not of stroke
J K K Vishram1, T Sehestedt2, T S G Sehested, S Lyngbaek2, J Jeppesen2, T W Hansen3, C Torp-Pedersen4, M H Olsen2
1Department of Internal Medicine, Glostrup Hospital and Research Center for Prevention and Health, Glostrup, Denmark, 2Department of Internal Medicine, Glostrup Hospital, Glostrup, Denmark, 3Department of Clinical Physiology and Nuclear Medicine, Hvidovre Hopsital, Hvidovre, Denmark, 4Department of Cardiology, Gentofte Hospital, Gentofte, Denmark
Topic: Risk factors and risk prediction
Purpose: The SCORE table is increasingly used for cardiovascular (CV) risk prediction in Europe. In the HeartScore programme it is possible to use the ratio between total cholesterol (TC) and high density lipoprotein cholesterol (HDL-C) instead of just TC. Whether it is possible to do the same in the SCORE table is uncertain. We investigated the impact of using TC/HDL-C instead of TC in the SCORE table.
Methods: In a population based sample of 2.109 subjects without CV disease or diabetes not receiving any CV, anti-diabetic or lipid-lowering treatment, aged 41, 51, 61, or 71 years we measured traditional CV risk factors in 1993–94. The SCORE and SCORE-HDL estimate were calculated using either TC or TC/HDL-C. CV death, fatal and non-fatal stroke, fatal and nonfatal ischemic heart disease (IHD) (fatal and non-fatal myocardial infarction or hospitalization for angina), as well as a composite CV endpoint (CEP) were assessed in 2006.
Results: Over a median follow-up of 13 years, the incidence of CV death, fatal and non-fatal stroke and IHD, as well as CEP amounted to 98, 55, 142 and 235 cases, respectively. Using SCORE-HDL instead of SCORE, the CV risk estimate increased in 18 subjects (0.9%) and decreased in 371 subjects (17.6%) (P>0.001). Subjects with a reduced CV risk estimate were older (26% vs. 12% were 71 years old, P>0.001), with higher TC (6.5±0.9vs. 6.0±1.1, P>0.001), higher HDL-C (1.66±0.4 vs. 1.41±0.4, P>0.001) and lower triglycerides (1.22±0.63 vs. 1.42±0.99, P>0.001). Including SCORE-HDL and SCORE in the same multiple Cox-regression analyses, their predictive values were compared. In these mutually adjusted models, SCORE-HDL predicted IHD significantly better than SCORE (hazard ratio pr 1 unit increase [HR=1.11 [1.03–1.18], P>0.01 vs. HR=1.02 [0.96–1.08], NS), whereas SCORE predicted stroke significantly better than SCORE-HDL (HR=1.11 [1.02–1.21], P>0.05 vs. HR=1.00 [0.90–1.11], NS). Both SCORE and SCORE-HDL predicted CV death equally well (HR=1.06 [1.01–1.12], P>0.05 and HR=1.07 [1.01–1.12], P>0.05, respectively). However, SCORE-HDL predicted CV death better than SCORE (HR=1.10 [1.02–1.12], P>0.05 vs. HR=1.05 [0.99–1.11], NS) in men, whereas SCORE predicted CV death better than SCORE-HDL (HR=1.50 [1.10–2.05], P>0.05 vs. HR = 0.89 [0.54–1.47], NS) in women.
Conclusion: SCORE-HDL CV risk estimate was significant lower in 17.6% of the subjects characterized by being older, with lower triglyceride, and higher TC and HDL-C. Using SCORE-HDL instead of SCORE, prediction of CV death was improved in men, but reduced in women. SCORE-HDL predicted IHD better, whereas SCORE predicted stroke better.
P63
Pulse pressure and subclinical atherosclerosis in young adults
E E Babes, A Ardelean, MI Popescu, VV Babes, M Rus, C Bustea
Faculty of Medicine, Oradea, Romania
Topic: Risk factors and risk prediction
Background: Endothelial dysfunction is an early event in atherosclerosis that precedes structural changes in the vascular wall. Carotid intimae-media thickness (CIMT) is a marker of structural atherosclerosis. These two indicators can be used to evaluate subclinical atherosclerosis.
The aim of the study was to evaluate subclinical atherosclerosis in a group of healthy young adults with cardiovascular risk factors and to evaluate how cardiovascular risk factors and particularly blood pressure and pulse pressure influence subclinical atherosclerosis.
Methods: We evaluated 102 young healthy adults (aged between 20–40 years, 58 men) with at least one cardiovascular risk factor. In all subjects we determined: smoking status, body mass index (BMI), blood pressure (BP), pulse pressure (PP), glicemia, lipid profile. Subclinical atherosclerosis was assessed with two indicators: endothelial dysfunction evaluated by determining brachial flow mediated dilation (FMD) and CIMT.
Results: FMD correlated directly with HDL colesterol (r=0,3, p>0,002) and inversely with age (r= − 0,26, p>0,01), BMI (r= − 0,2, p>0,05), glicemia (r= − 0,19, p=0,05), CIMT (r= − 0,89, p>0,0001), systolic BP (r= − 0,85, p>0,0001) and PP (r= − 0,88, p>0,0001). In multiple regression analysis only CIMT (r= − 0,89, p=0,02) and PP (r= –0,88, p=0,04) remained independent predictors for FMD. CIMT correlated directly with age (r=0,27, p=0,005), BMI (r=0,3, p=0,002), LDL (r=0,22, p=0,02), PP (r=0,95, p>0,0001), systolic BP (r=0,94, p>0,0001) and inversely with FMD (r= − 0,89, p>0,0001), and HDL (r= − 0,41, p>0,0001). In multiple regression analysis only FMD (r= − 0,89, p=0,002) and PP (r=0,95, p=0,016) remained independent predictors for IMT. PP was significantly directly associated with IMT and inversely with FMD even after adjustment for age, BP, cholesterol, BMI, glicemia. There is a direct correlation of subclinical atherosclerosis with systolic BP which did not persist in multiple regression. The significant inverse association between IMT and FMD, implies that a preserved endothelial function will delay the onset of atherosclerosis.
Conclusions: PP is an independent risk factor for subclinical atherosclerosis in young adults. PP could predispose to atherosclerosis by influencing endothelial function and nitric oxide release or representing an injury to the arterial wall. Further investigations are requested to evaluate whether a treatment for reducing PP could better prevent atherosclerosis and cardiovascular disease.
P64
Long term effect of a structured postcare to cholesterol levels after stationary rehabilitation
M Wonisch1, B Eder1, D Brandt1, R Mueller2
1Center of Cardiac Rehabilitation, St. Radegund/Graz, Austria, 2PV, Hauptstelle Chefärztlicher Bereich, Vienna, Austria
Topic: Risk factors and risk prediction
Background: Initiation of a long-term improvement of cardiac risk factors is one of the major aims of cardiac rehabilitation/secondary prevention program. Positive short-term effects of a 4-weeks residential rehabilitation program have been published. However, an ongoing effect has not clearly been shown.
Methods and Participants: A health guide collected data in terms of cardiac risk factorsat admission and release after 3 until 4 week stationary rehabilitation program. At first the goal of each parameter were defined in corporation with the physician. In the period of 3 months the parameters in the health guide were completed. 12 months later the health guides were returned to the rehabilitation centre. In the prospective study 2664 patients (71.8% men, age: MV=62.94 years, SD=9.96; 28% women, MV=67.59 years, SD=9.53) with coronary heart disease (CHD) were included.
Results: Total cholesterol (TC) and LDL-cholesterol (LDL) improved during the 3 until 4 weeks of cardiac rehabilitation (TC: 173±41vs. 154±34mg/dl, p>0.05; LDL: 112±36 vs. 93±28 mg/dl, p>0.05). The individual goals were reached in 83% and 66%, respectively. After one year we found an increase in TC (179±37mg/dl) and LDL (102±31mg/dl), but LDL was still significantly lower than ad admission.
In the last 8 years we found an improvement of TC and LDL at release and after 12 months, i.e. in the investigation years 2005 to 2007 TC and LDL were significantly lower than in the years 2000 to 2004 (Tab.).
Conclusion: The Health guide resulted in an improved long-term effects of cardiac risk factors. It is a cost-effective intervention and can help in the guidance of the patients. A better management regarding cholesterol was shown in the last 8 years.
Total- and LDL-cholesterol during 1 year
2000
2001
2002
2003
2004
2005
2006
2007
n=161
n=368
n=383
n=425
n=361
n=349
n=376
n=220
Total Cholesterol (mg/dl)
Admission
180±42
178±44
180±42
178±44
174±40
166±34
162±36
166±37
Release
164±30
158±35
162±36
160±36
157±34
145±30
142±28
140±28
12 months
188±36
187±40
184±37
181±36
177±33
171±35
168±35
168±33
LDL-Cholesterin (mg/dl)
Admission
113±37
115±38
114±36
112±36
107±33
104±30
101±32
106±32
Release
98±26
98±30
98±29
97±30
91±26
87±28
85±25
84±25
12 monts
107±31
110±32
107±32
104±33
97±28
96±29
94±28
92±27
P65
High level of low density lipoprotein cholesterol in adult children of patients with premature coronary heart disease: relation to own and parental characteristics
MV Konnov, LM Dobordzhginidze, AD Deev, NA NikolaiGratsiansky
Institute of Physico-Chemical Medicine, Moscow, Russian Federation
Topic: Risk factors and risk prediction
High level of low density lipoprotein cholesterol (LDL-C) can be a critical contributor to elevated risk of coronary events in children of patients with premature coronary heart disease (CHD).
Purpose: To elucidate associations between high LDL-C in adult children of patients with premature (onset >55 years, men; >60 years, women) CHD and some their own and parental characteristics.
Methods: We examined members of 179 families. There were 171 parents-probands (65.2% men) aged 40–63 years, their 113 consorts (24.1% men) aged 37–67 years and 215 their own children (50.7% men) aged 18–37 years. Characteristics studied included alcohol consumption, tobacco smoking, oral contraceptive use, education level, presence of menstruation, height, body mass index (BMI), waist circumference (WC), heart rate, systolic and diastolic blood pressure (DBP), total and high density lipoprotein cholesterol, LDL-C, triglycerides, basal serum glucose, impaired glucose tolerance, diabetes mellitus (ADA criteria), arterial pre- and hypertension (NHBPEP-4 or JNC-7 depending on age), metabolic syndrome (MS, IDF criteria), use of triglyceride and LDL-C lowering drugs. High LDL-C was defined as 3.36 mmol/l. Its predictors were selected by sex and age adjusted logistic regression analysis.
Results: High LDL-C was found in 84/215 (39.1%) of children. Characteristics related to the presence of high LDL-C with p>0.1 in univariate analysis (own smoking, DBP, BMI; LDL-C of proband; BMI, WC, LDL-C and MS of non-proband) were included into stepwise regression procedure. Characteristics independently associated with high LDL-C were: LDL-C of aparent -proband; BMI and LDL-C of a parent-non-proband own non-smoking (table).
Conclusions: In this group of adult children with parental premature CHD LDL-C level of a parent with CHD was in dependently associated with high LDL-C. However characteristics of parents without CHD (LDL-C and BMI) were also independently related to high LDL-C in their adult children.
Results of multivariate analysis
OR
95% CI
p
LDL-C of proband, mmol/l
1.46
1.12–1.91
0.005
LDL-C of non-proband, tertile 3 ([=4.42 mmol/l] vs tertile 1 [=3.71 mmol/l])
4.26
1.59–11.4
0.004
BMI of non-proband, tertile 3 ([=29.4 kg/m2] vs tertile 1 [=25.5 kg/m2])
3.58
1.34–9.55
0.011
Own smoking vs non-smoking
0.46
0.24–0.88
0.018
(W) P66: Influence of cardiovascular risk factors on age at presentation with the first acute coronary syndrome
J D Brito, J Ferreira, P Carmo, C Aguiar, M Trabulo, J Aniceto Silva
Hospital Santa Cruz, Carnaxide, Portugal
Topic: Risk factors and risk prediction
Purpose: This study aimed to investigate the effect of traditional risk factors hypertension, hypercholesterolemia, smoking and diabetes mellitus- plus an emerging risk factor obesity- on the age at presentation with acute coronary syndrome (ACS).
Methods: We studied 428 patients without history of cardiovascular (CV) disease who were admitted in our hospital with ACS. The effect of risk factors on age at presentation with ACS was evaluated by Student's t-test and linear regression.
Results: Mean age was higher in women (63.3±12.6 vs 59.2±11.6 in men; p=0.003) and in non-ST-elevation ACS (63.0±12.6 vs 57.5±11.7 in STEMI; p>0.001). Hypertension was present in 55.8% of patients, 52.1% had hypercholesterolemia, 38.3% were smokers at the time of admission, 15.4% had diabetes and 17.9% had body mass index>30kg/m2. Smoking lowered the age of presentation by 17.9 years in women (CI95%: 12.7–23.2; p>0,001) and 10.1 years in men (CI95%: 7.8–12.3; p>0.001). Hypertension delayed appearance of ACS by 12.4 years in women (CI95%: 7.4–17.3; p>0.001) and 4.9 years in men (CI95%: 2.5–7.3; p>0.001). In men, ACS presented 3.4 years earlier (CI95%: 0.37–6.4; p=0.028) in those with BMI>30 and 3.1 years earlier (CI95%: 0.6–5.5; p=0.015) in those with hypercholesterolemia. Diabetes had no effect on age of presentation with ACS. Table 1 depicts the risk factors which remained independent predictors of age of presentation with ACS. In the linear regression model, there was a relationship between the age of presentation with ACS and regular use of angiotensin converting enzyme (ACE) inhibitors (beta coefficient = 0.126, p=0.017).
Conclusions: Age of presentation with ACS is significantly affected by CV risk factors, especially in women. The strongest association occurs with smoking. Hypertensive patients are older at presentation with ACS, which may be partially explained by regular use of ACE inhibitors.
Men (R=0.502)
Women (R=0.615)
Risk factor
β coefficient
p
Risk factor
β coefficient
p
Smoking
–0.412
>0.001
Smoking
–0.473
>0.001
Hypertension
0.191
>0.001
Hypertension
0.235
0.022
BMI>30 kg/m2
–0.144
0.004
Hypercholesterolemia
–0.115
0.020
Independent predictors of age of presentation with ACS.
P67
Admission glycaemia beats glycaemia variation in predicting post myocardial infarction outcome (only) in non-diabetics
S Monteiro, C Lourenco, R Teixeira, R Batista, E Jorge, P Lazaro, P Monteiro, L Providencia
University Hospitals of Coimbra, Coimbra, Portugal
Topic: Risk factors and risk prediction
Introduction: Hyperglycaemia during acute myocardial infarction (AMI) is associated with bad prognosis. There are several parameters to access glucose metabolic control, but is there one which is clearly better?
Aim: To evaluate the impact of admission glycaemia (AG) and magnitude of glycaemia variation (MGV) on the prognosis of diabetic and non-diabetic AMI patients and to identify independent predictors of mortality post-AMI.
Population and methods: 1053 consecutive AMI patients, diabetics (n=348) and non-diabetics (n=705). Both groups were then divided, according to the quartiles of AG (Q1 >5.83, Q2 5.83–7.44, Q3 7.44–10.33, Q4 =10.33 mmol/l) and MGV (Q1 >0.78, Q2 0.78–1.67, Q3 1.67–3.33, Q4 =3.33 mmol/l). Patients were followed during 1-year. MGV was defined as the difference between AG and the lowest fasting glycaemia during hospitalization.
Results: In diabetics there were no differences in mortality among AG and MGV quartiles. In non-diabetics, higher AG was associated with higher in-hospital and 1-year mortality, while higher MGV was only significantly related to 1-year mortality (Table 1).
After multivariate regression analysis, AG=5.83 mmol/l (OR 8,91; IC 1,20–66,30), age = 72 years (OR 3,58; IC 1,79–7,20), Killip class < 1 (OR 3,49; IC 1,87–6,52) and Troponin I = 6.0 ng/ml (OR 2,68; IC 1,28–5,58) were independent predictors of in-hospital mortality and Q4 versus Q1 of MGV (OR 2,50; IC: 1,22–5,12) was an independent predictor of 1-year mortality.
Conclusion: In our AMI population, AG was a short and long-term bad prognosis marker, while MGV impacted only in long-term prognosis, especially in non-diabetics. These data suggest that AG is a more complete mortality predictor than MGV.
P68
Gender differences in the prevalence of cardiovascular disease risk factors and corresponding 10-year cardiovascular risk, among subjects with the metabolic syndrome
T Chimonas1, V Athyros2, E Ganotakis3, V Nicolaou4, D Panagiotakos5, E Elisaf1
1University of Ioannina Medical School, Ioannina, Greece, 2Aristotle University of Thessaloniki, Thessaloniki, Greece, 3University of Crete, Heraklion, Greece, 4Red Cross Hospital, Athens, Greece, 5Harokopio University, Athens, Greece
Topic: Risk factors and risk prediction
Purpose: Cardiovascular disease (CVD) is more prevalent in men than in women; however, the clustering of CVD risk factors among people with metabolic syndrome (MS) has rarely been investigated between genders. Thus, we sought to evaluate gender differences in CVD risk factor prevalence and the estimated 10-year CVD risk in a Greek population.
Methods: A random sample of 824 male and 1199 female subjects with MS (NCEP ATPIII), but without diabetes mellitus or established CVD, was selected from several Greek areas. Blood lipids, blood pressure (BP), fasting glucose (FG), waist circumference and other biochemical markers were measured. Ten-year risk estimates for fatal CVD were calculated using the ESC SCORE and Framingham models.
Results: Women with the MS were older than men (57.6±10.3 vs 55.7±11.1 years, p>0.001). Elevated BP levels (>130/85 mmHg) and hypertriglyceridemia (=150 mg/dl) were more common in men than women (90.0 vs. 85.9% and 86.8 vs. 75.2%, respectively; p>0.001), whereas low HDL-C (>40 mg/dl in men and >50 mg/dl in women) and abdominal obesity were more common in women (59.1% vs. 65.1% and 83.6% vs. 97.0%, p>0.001). The total number of metabolic criteria was equally distributed between the sexes, with 35.0% of men vs 32.7% of women showing any 3 criteria, 39.4 vs 39.9% any 4 criteria and 25.6 vs 27.4% exhibiting all 5 defining criteria (p not significant for all comparisons). The 10-year risk for fatal CVD events was almost threefold higher in men than in women (8.0±8.7% vs 3.0±3.8%, p>0.001 using SCORE and 8.6±8.1% vs 3.6±4.2%, p>0.001 using Framingham).
Conclusions: The MS is influenced by different factors in men and women, with men being at significantly higher risk for CVD. This information can be used for planning a better population-wide strategy for the prevention and treatment of the MS in both sexes.
P69
Which risk factors influence coronary events in Japanese high-risk hypertensive patients?: A subanalysis of the CASE-J (Candesartan Antihypertensive Survival Evaluation in Japan) trial
K Ueshima1, S Yasuno1, K Oba1, A Fujimoto1, T Ogihara2, T Saruta3, K Nakao
1EBM Research Center, Kyoto University Graduate School of Medicine, Kyoto, Japan, 2Osaka General Medical Center, Osaka, Japan, 3Keio University, Tokyo, Japan
Topic: Risk factors and risk prediction
CASE-J trial group
Objectives: The CASE-J trial had been designed to compare the long-term effects of the angiotensin II receptor blocker candesartan and the calcium channel blocker amlodipine on the incidence of cardiovascular events in Japanese high-risk hypertensive patients. The present study was aimed to evaluate the influence of various risks on coronary events in Japanese high-risk hypertensive patients as a subanalysis of the CASE-J trial.
Methods: We considered the CASE-J trial as an observational study irrespective of allocated drugs, and focused on the possible risk factors which influence coronary events in all of the 4,703 enrolled patients. We examined the adjusted hazard ratios (HRs) of various baseline possible risk factors including sex, age, allocated drugs, BMI, systolic/diastolic blood pressure (SBP/DBP), diabetes mellitus (DM), hyperlipidemia, smoking, left ventricular hypertrophy, previous cerebrovascular events, chronic kidney disease (CKD), and previous ischemic heart disease (IHD), on the coronary events (sudden death, angina pectoris, and acute myocardial infarction). Furthermore, patients were classified into two groups stratified according to presence or absence of previous IHD at baseline, and the same analyses were performed in each group.
Results: Mean SBP/DBP were strictly controlled less than 140/80 mmHg. During 3.2 years of follow-up, the coronary events occurred in 83 (1.8%) patients, and were significantly associated with previous IHD, DM, male, CKD, and low DBP. HRs of these factors were previous IHD (HR, 3.89: CI, 2.40–6.31, p>0.001), DM (3.10:1.90–5.04, p>0.001), male sex (1.81:1.06–3.10, p=0.031), CKD (1.60:1.01–2.54, p=0.046), and low DBP (1.36 per 10 mmHg decrease:1.09–1.68, p=0.006), respectively. In 4,107 patients without previous IHD at baseline, the significant predictors were DM (4.88: 2.35–10.16, p>0.001), hyperlipidemia (2.67:1.47–4.85, p=0.001), and DBP (1.39 per 10 mmHg decrease:1.05–2.18, p=0.020), while male sex (3.05:1.15–14.63, p=0.025), CKD (2.44:1.16–5.13, p=0.018), and DM (2.15:1.06–4.38, p=0.035) were significantly associated with the risk of coronary events in 596 patients with previous IHD at baseline.
Conclusions: Although there are some differences in the risk factors between the primary and secondary prevention in Japanese high-risk hypertensive patients, DM is the important factor in both preventions. Thus, comprehensive risk management including DM, CKD and hyperlipidemia should be done for preventing the coronary events in addition to blood pressure control in hypertensive patients.
P70
Clinical value of walk-test and nt-proBNP in asymptomatic aortic stenosis.
M Gomez Perez1, L Molina1, M Cladellas1, A Ramirez1, J Vila2, C Soler1, H Sanz2, J Bruguera1
1Hospital Del Mar, Barcelona, Spain, 2Institut Municipal Investigacio Medica, Barcelona, Spain
Topic: Risk factors and risk prediction
Introduction: N-terminal-pro-Brain Natriuretic Peptide (NT-proBNP) could be a prognostic marker in some valve diseases as aortic stenosis (AE). 6-minutes-walk test (WT) provides information about functional class in this patients
Objectives: To assess prognostic value of NT-proBNP and its correlation with 6-minutes-walk-test in patients with asymptomatic aortic stenosis.
Material and methods: Prospective study of 180 patients with severe aortic stenosis (Mean Aortic Valve Area 0.8+/–0.3 cm2, Mean gradient 45+/–13 mmHg, Ejection Fraction 61+/– 14%). Mean age: 77 (4.3) years. Echocardiographic studies were performed with Vivid 7 (General Electric). We considered clinical event as: symptoms related to AE (angina, heart failure or syncope), surgical treatment or mortality. In all patients were determined NT-proBNP in the sera and performed a 6 min-Walk-Test.
Results: After a mean follow-up of 18 (4.7) months, 59 patients suffered some clinical event. Mean NT-proBNP was 1988(4125) pg/mL (21–23304). Patients with events in follow-up presentedlower exercise capacity (131 vs 238 m, p=0.001). NT-proBNP correlated with WT (r=–0.78, p>0.01). The best NT-proBNP Cut-point value for prediction of events was 241 (95%CI: 100–1412) pg/mL. Higher levels of NT-proBNP and poor distance in WT, predicted events in the follow-up (p>0.001)
Conclusion: NT-proBNP and Walk-Test provides a prognostic information in patients with asymptomatic severe aortic stenosis
P71
Endothelial dysfunction and carotid lesions are strong predictors of clinical events in patients with early stages of atherosclerosis: a 24 months follow-up study
MI Popescu, M Rus, AI Ardelean, EE Babes, BB Babes, CM Bustea
University of Medicine and Pharmacy, Oradea, Romania
Topic: Risk factors and risk prediction
Grant CNCSIS
Background: Recent findings suggest that endothelial vasodilator dysfunction and carotid intima-media thickening (IMT) may represent good indicators of sub clinical cardiovascular diseases. The purpose of this study was to investigate whether the vasodilator response to brachial artery and the presence of carotid lesions may have a prognostic significance in patients with early stages of atherosclerosis.
Methods and Results: Vascular echography was performed to analyze flow mediated vasodilatation (FMD) at the brachial artery and IMT of carotid arteries in 84 asymptomatic subjects. At baseline we evaluated all the established traditional cardiovascular risk factors, including early age, male gender, obesity, hypertension, diabetes, smoking habit, family history of CAD and plasma lipids. Effort or unstable angina, acute myocardial infarction, served as outcome variables over a follow-up period of 24 months. Brachial FMD was correlated inversely with carotid IMT (p=.003), systolic blood pressure (p=.0001) and age (p=.0001). IMT was positively correlated with systolic blood pressure (p=.0001), waist circumference (p=.004) and age (p=.01). At the end of the follow-up cardiovascular events were registered in the 29% of patients and in a multivariate analysis, including all the variables evaluated at baseline, male gender (p=.001), the presence of baseline carotid lesions (p=.01) and FMD below the median (p=.03) were the only variables predictive of clinical events.
Conclusions: In the present study, endothelial dysfunction and carotid lesions significantly increased the risk of vascular events in asymptomatic subjects with early stages of atherosclerosis. Assessment of systemic vasoreactivity and carotid IMT evaluation may provide in this category of patients important prognostic information in addition to those derived from traditional established cardiovascular risk factors.
P72
The relationship between overweight and obesity and diabetes. The results of Multi-Center All-Polish Health Survey (WOBASZ)
M Polakowska, W Piotrowski
Institute of Cardiology, Warsaw, Poland
Topic: Risk factors and risk prediction
Purpose: The relationship between overweight and obesity and diabetes occurrence.
Methods: Data comes from WOBASZ research carried on from 2003 to 2005 years. To analysis includes 14169 people (6666 men,7503 women). BMI index and waist circumference were used to determine overweight and obesity. The use of logistic regression model allowed estimation of influence of prognostic value of BMI and waist (AUC) on diabetes occurrence. The BMI and waist circumference cut-off points, that allow classification to the proper diabetes group with biggest accuracy and specificity, were determined using characteristic-operational curves and Youden criteria.
Results: The diabetes frequency grew in conjunction with BMI and came to 3% by men and 2% by women, within normal weight group, 17% and 16% within obesity group and also according to waist circumference 5% by men and 2% by women with correct waist and 16% and 13% with incorrect waist. Together with grow of BMI and waist, grows the diabetes percentage to 18% by men and 17% by women. The risk of diabetes was higher by women than men, both by incorrect BMI and waist circumference. The waist circumference has better prediction on diabetes occurrence. AUC by men and women was about 10% bigger by waist than BMI index. The best cut-off points for BMI are: 28,6 by men and 28,1 by women, and for waist: 98,9 and 91,0 cm.
Conclusions: The diabetes frequency grows with incorrect parameters describing overweight and obesity. Waist circumference has bigger predictive value of diabetes occurrence. BMI and waist circumference has bigger predictive value by women.
(W) P73: Preventing harm associated with warfarin: identification of risk factors in the community
B Diug1, J Lowthian1, S Evans1, E Maxwell2, M Dooley3, A Street3, P Cameron1, J Mcneil1
1Monash University, Melbourne, Australia, 2Melbourne Pathology, Melbourne, Australia, 3The Alfred Hospital, Melbourne, Australia
Topic: Risk factors and risk prediction
Purpose: This study aimed to develop a comprehensive risk profile of patients on warfarin therapy who were at increased risk of bleeding within the community.
Methods: A case control study was conducted with patients recruited by a metropolitan pathology provider. An INR = 6.0 was used as a surrogate for increased bleeding risk. Patient interviews investigated potential predisposing factors including demographics, co-morbidities, diet, medication, and warfarin knowledge. Standardised measures evaluated cognition, mood, social support, health literacy, functional independence and adherence.
Results: Total of 486 patients: 157 cases (mean age 75.4 yrs, range 25–96) and 329 controls (mean age 75.5 yrs, range 36–92). Median duration of treatment was 4 years (0.5–31) in cases and 5 years (0.6–45). Atrial fibrillation was the most common primary indication. Multivariate logistic regression analysis was conducted (see Table).
Conclusion: Our findings demonstrate that there is a strong association between: psychosocial deficits - inadequate functional health literacy, depression, mild cognitive impairment; specific medications and co-morbid illnesses and increased bleeding risk. We recommend that patients on warfarin be reviewed regularly in a holistic manner to identify potential underlying characteristics that have the potential to prevent adverse events.
Variables
OR (95% CI)
p-value
Inadequate functional health literacy
4.1(1.2–3.5)
0.0001
Depression
2.7(1.6–4.4)
0.0001
Mild cognitive impairment
1.9(1.1–3.5)
0.02
Working status - Retired
0.2(0.1–0.4)
0.0001
Primary Indication - Atrial Fibrillation
0.3(0.2–0.6)
0.0001
Poor visual acuity
2.8(1.7–4.9)
0.0001
Concomitant Antibiotics
7.1(2.9–16.7)
0.0001
Concomitant Prednisolone
3.2(1.3–7.6)
0.01
History of Diabetes
1.7(1.1–3.1)
0.05
Concomitant medication - >5
reference
n 5–10
2.1(1.2–3.8)
0.04
n =10
2.7(1.3–5.6)
0.01
P74
Predictor factors of valvular calcification in chronic kidney disease
A F Ana FranciscaSimoes Marques Assuncao Caetano, S Barra, A Botelho, N Oliveira, P Mota, A Carreira, A Leitao Marques
Hospital Center of Coimbra, Coimbra, Portugal
Topic: Risk factors and risk prediction
Introduction: Valvular calcification (VC) is common in patients (P) with chronic kidney disease (CKD). Recent data have shown an unfavourable prognosis associated with their presence, so it is important to understand its pathophysiology.
Purpose: Assess the prevalence of VC in P with CKD and identify key predictors of VC.
Population and methods: Retrospective study including 84 P (53,6% male, mean age 67,1±15,8 years) with the diagnosis of CKD (K-DOQUI guidelines classification-NKF: 86,9% stage 5; 8,3% stage 4 and 4,8% stage 3 of CKD) admitted to a Nephrology department, between June 2007 and May 2009. 42,9% were on maintenance haemodialysis (HD), mean time on dialysis 29±25 months. In this population, P with VC had higher incidence of atrial fibrillation (p=0,02), aortic stenosis (p>0,001) and mitral regurgitation (p=0,079). Clinical, analytical and ecochardiographic parameters were compared.
Results: 44% of P had VC − 7,1% of mitral valve; 17,9% of aortic valve and 19% of both valves. VC was found in 39,6% of P without maintenance HD and on 50% of P on HD (p=n.s.). It was found a statistically significant association between the following factors and the presence of VC: older age (60,2 Vs 74,5; p<0,001); higher blood levels of calcium (>2,1 Vs = 2,1; p=0,022), alkaline phosphatase (=136 Vs < 136; p=0,018) and parathyroid hormone (in the group of P on HD: =150 Vs <150; p=0,031). P treated with beta-blocker had lower incidence of VC (p=0,008). There was no association of VC with the following factors: sex, stage of CKD, length of HD; treatment with sevelamer Vs calcium carbonate; cholesterol levels (total; HDL or LDL); triglycerides; serum phosphorus; phosphorus-calcium product and C-reactive protein.
Conclusions: VC is a frequent echocardiographic finding in P with CKD, being present even before the onset of HD. In a significant proportion of cases there was simultaneous involvement of the mitral and the aortic valve. VC is associated with the presence of valvular heart disease and arrhythmias. High levels of serum calcium and pharathyroid hormone are present in P with VC. They may be associated with the deposition of calcium in the heart, suggesting that an optimization of phosphorus-calcium metabolism may slow down its development and complications. Regarding the “protector” effect of the beta-blocker found in this study, it should be questioned in view of the limited number of patients and the controversial pleotropic effect of this class of drugs.
P75
Assessment of arterial damage in metabolic syndrome
V Aursulesei, A Cozma, MD Datcu
University of Medicine and Pharmacy “Gr. T. Popa”, Iasi, Romania
Topic: Risk factors and risk prediction
Purpose: To evaluate the extent and features of the arterial damage in metabolic syndrome using several predictors of clinical outcome: pulse wave velocity (PWV), intima media thickness (IMT), left ventricular hypertrophy (LVH), ankle-brachial index (ABI) and creatinine clearance.
Material and methods: We studied 153 consecutive patients with metabolic syndrome and 89 control subjects, matched for mean age (59.7 ± 2.1 vs 61.7 ± 1.3 years) and sex distribution (78.7% vs 76.1% males). The following noninvasive markers were used for evaluation: carotid-femoral PWV (Complior), carotidian IMT, LVH defined by index of mass (LVMI) and geometric patterns, ABI and creatinine clearance.
Results: In patients with metabolic syndrome the arterial damage is extended in different territories in 73.8%. There are significantly differences for all parameters between the metabolic syndrome group and control subjects: PWV = 15.9 ± 2.3 vs 9.00 ± 1.38 m/s (p = 0.003), carotidian IMT = 0.98 vs 0.76 mm (p = 0.04), ABI = 0.69 vs 0.87 (p = 0.002), LVMI = 123.1 vs 104.4 g/m2 (p>0.05), creatinine clearance = 63.1 vs 95.6 ml/min (p>0.05). In the metabolic syndrome group PWV is strongly related with LVMI and concentric hypertrophy, irrespective the gender (p>0.05) and carotidian IMT in males (r = 0.38, p = 0.01). Increased values of PWV and decreased ABI are related with number of traits of metabolic syndrome after adjusting for confounders/other cardiovascular risk factors (ANCOVA). The creatinine clearance is only related with the mean value of ABI (r = 0.32, p>0.01). Postprandial hyperglicaemia and systolic blood pressure seem to be independent predictors for both PWV and ABI, while HDL-cholesterol is independently related with carotidian IMT in metabolic syndrome.
Conclusion: Our results suggest that the clustering of MS components may interact to synergistically affect the degree and extent of arterial damage. PWV and ABI are valiant parameters for a complete assessment of arterial damage in metabolic syndrome. The combination of these measurements might be of stronger relevance for the clinical outcome.
P76
Predictors of cardiovascular risk 24 months after a Myocardial Infarction: the importance of clinical and analytical parameters
S N C Barra, P Gomes, J Silva, R Providencia, L Seca, A L Marques
Hospital Center of Coimbra, Coimbra, Portugal
Topic: Risk factors and risk prediction
Purpose: Risk factors for acute myocardial infarction (MI) mortality have been thoroughly studied. Analysis of the significance of several clinical and analytical parameters as predictors of cardiovascular (CV) risk in the 24 months following a MI could allow a more aggressive therapeutic approach in those patients most at risk.
Methods: 404 consecutive patients (68.1±13.7 years old; 63.4% males; 37.4% with Diabetes Mellitus) discharged after being admitted for a MI. Patients were followed for 2 years. Admission glycaemia, HbA1c, creatinine, haemoglobin [Hb] and C Reactive Protein (PCR); maximum troponin levels (MaxTrop); glomerular filtration rate (GFR) by MDRD formula; maximum Killip Class (mKC), and coronary artery disease (CAD) extension were measured to assess their usefulness as predictor(s) of CV risk (symbolized by mortality risk, risk for reinfarction, recurrent angina, and admission for decompensated heart failure [DHF]).
Results: See Abstract Table.
Conclusions: Following discharge after a MI infarction, admission glycaemia, Hb, creatinine, GFR and PCR showed high predictive value for CV events. A more extensive CAD and the occurrence of acute heart failure were also strong predictors of events during follow-up. These data reinforce the need to monitor such risk markers in at risk patients.
Predictors of cardiovascular risk.
Mortality
Reinfaretion
Recurrent angina
DHF
Glycemia = 7.7mmol/L
28.9% vs 11.8%; p=0.001; OR 3.06
p=n.s.
P=n.s.
25.7% vs 10.9%; p=0.002; OR 2.83
Hb<12 mg/dL
46.2% vs 13.3% p>0.001; OR 5.59
27.8% vs 10.7%; p=0.001; OR 3.20
P=n.s.
28.8% vs 14.7%; p=0.017; OR 2.35
HbA1c < 7%
p=n.s.
p=n.s.
P=n.s.
p=n.s.
Creatinine <120çmol/L
40% vs 13.9%; p>0.001; OR 4.14
29.5% vs 9.8%; p>0.001; OR 3.87
P=n.s.
37,7% vs 11.4%; p>0.001; OR 15
GFR<60mL/min/m2
35.2% vs 12.2%; p>0.001; OR 3.9
26.7% vs 8.0%; p>0.001; OR 4.16
P=n.s.
31.8% vs 10.4% p>0.001; OR 4.02
PCR < 3 mg/dL
p=n.s.
44.4% vs 10.5%; p=0.001; OR 6.8
58.8% vs 21.1%; p=0.003; OR 5.36
p=n.s.
MaxTrop < 42 ng/mL (MaxTrop50)
p=n.s.
p=n.s.
P=n.s.
p=n.s.
mKC < 1
34.3% vs 14.9% p=0.001; OR 3.0
p=n.s.
P=n.s.
32.4% vs 12.3%; p>0.001; OR 3.41
=2 affected coronary segments
16.5% vs 6.5%; p=0.027; OR 2.84
14.4% vs 4.3%; p=0.014; OR 3.74
32.8% vs 20.7%; p=0.048; OR 1.88
p=n.s.
P77
The association between social class and mortality after major elective cardiovascular surgery
Akdeniz University, Faculty of Medicine, Antalya, Turkey
Topic: Public health
Background: Several studies have shown the powerful association between a person's socioeconomic status and mortality. Also studies have shown that income inequality is correlated with overall mortality. This study was set up to measure the association between social class and early and late mortality after major elective cardiovascular surgery
Methods: This cohort was conduct among patients seen in cardiovascular surgery clinics. Each of 521 patients who accepted to participate in the study was face to face interviewed to complete a personal data form. Socioeconomic variables of the patients were entered into personal data forms and the participants were divided into eight social classes according to these variables: blue collars, white collars, qualified and unqualified workers, high and petty bourgeois, self-employed and unemployed persons. In 30 days mortality was accepted early mortality.
Results: The mean age was 58,4 ±14,1 years and the number of the female and male participants were 135 (25, 9%) and 386 (74, 1%) respectively. Of the patients, 354 (68, 0 %) had coronary bypass, 93 (17, 9 %) had valve replacement, 13 (2.5%) had cardiac transplantation, 47 (9, 0 %) had major vascular by-pass and 14 (2, 6 %) had unruptured abdominal aorta aneurysm.
The patients of blue collars, unqualified workers were more likely to die after major elective cardiovascular surgery (OR 2.02, p trend 0,038).
Conclusions: Disadvantaged people seem particularly vulnerable to mortality after major cardiovascular surgery. Factors related to social class are likely to influence poor prognosis after cardiovascular surgery.
P78
The resting heart rate strongly predicts all-cause death among middle aged women and men from a general population
G Nilsson, A Rosenblad, J Leppert
Center for Clinical Research, Västerås, Sweden
Topic: Cardiovascular epidemiology
The resting heart rate (HR) has been found to predict all-cause death among middle aged healthy men. Corresponding data for women are lacking. We report the sex specific predictive ability of heart rate for all-cause mortality among middle aged women and men.
Material and methods: Women and men from the county of Vastmanland, Sweden (260000 inhabitants) were invited to a health survey at the age of 40 and 50 during the years 1990–99 (participation rate 48%). HR was measured as radial pulse rate. Relative risk (RR) for all-cause death was computed by Cox regression (follow-up until August 17, 2008).
Results: Among 15965 men and 17250 women, 751 men and 537 women died during follow-up (deaths per person-year: men 0.34%, women 0.22%). The mean HR was 69 beats per minute (bpm) for men and 71 for women. The RR (95% confidence interval) for 1 bpm increase of HR was 1.034 (1.027–1.041); p>0.001 for men and 1.029 (1.021–1.038); p>0.001 for women. Multivariable adjustment (prevalent smoking, diabetes, hypertension, myocardial infarction, stroke as well as cholesterol, BMI, systolic blood pressure and age cohort) attenuated the RR; men 1.023 (1.016–1.031); p>0.001 and women 1.021(1.013–1.030); p>0.001.
The RR for people within the upper quartile (pulse rate>76) compared with those within the lowest quartile (pulse rate >65) was 2.3 (1.9–2.8) for men and 1.9 (1.5–2.4) for women.
Conclusion: HR is a strong predictor of all-cause death among middle-aged women and men.
P79
Comparative study between increased waist circumference and body mass index for deterioration on cardiac structure and function in male Korean rural population: Atherosclerosis RIsk of Rural Area iN Korea General Population
JY Kim, JK Sung, YJ Yoon, SH Lee, JH Yoon, KH Choe, JW Lee
Wonju College of Medicine, Yonsei University, Division of Cardiology, Department of Internal Medicine, Wonju, Republic of Korea
Topic: Cardiovascular epidemiology
ARIRANG Cohort study group
Purpose: Increased abdominal circumference and body mass index (BMI) are independent cardiovascular risk factor and related preclinical changes on cardiac structure and function. However, there is no comparative study between BMI and waist circumference for deterioration on the functional and structural changes of heart. Therefore, we evaluated what is more predictable component for deterioration on cardiac geometry and function in male Korean rural population.
Methods: We analyzed consecutive 1,462 male aged 4070 years from ARIRANG study. Abdominal circumference and BMI were categorized in tertile and three by three tables were made. All subjects were measured cardiac geometry, systolic and diastolic function by echocardiography. To evaluate the effect of each table on LV mass, LV chamber size, LA size, and diastolic dysfunction, we used logistic regression method for each table with adjust for age, HOMA-IR, systolic blood pressure, and pulse rate.
Results: Mean age was 53.5 ± 7.8. Both increasing BMI and waist circumference were significantly associated with increased LV end-diastolic dimension, LV mass index, LV mass to height 2.7, LA volume index and deceleration time of E wave (p>0.01). However, relative wall thickness and LV ejection fraction were showed no statistical significant in increased BMI and waist.
In logistic regression model, increasing BMI was significant predictor of high index LV mass to height2.7 than abdominal circumference. Increasing abdominal circumference was significant predictor of increased LV end-diastolic dimension, LA volume and diastolic dysfunction than BMI (Fig. 1).
Conclusion: Central obesity is more associated with diastolic dysfunction, LA and LV enlargements than BMI. BMI is more important factor for LV hypertrophy than waist circumference.
Comparison of odd ratio between increase
P80
Relative importance of risk factors for cardiovascular disease incidence and mortality: a prospective cohort study
J Igland1, SE Vollset, O Nygard2, HK Gjessing1, GS Tell1
1University of Bergen, Department of Public Health and Primary Health Care, Bergen, Norway, 2University of Bergen, Institute of Medicine, Bergen, Norway
Topic: Cardiovascular epidemiology
Lifestyle Epidemiology
Purpose: The purpose of the study was to rank established cardiovascular risk factors according to importance in order to identify the strongest predictors of cardiovascular incidence and mortality. Of particular interest was to examine whether the ranking of risk factors was different for men and women and for different age groups.
Method: The Hordaland Homocysteine Study, a population-based cohort study, was conducted in 1992–93 in Hordaland County, Western Norway. The 18,044 participants (52.5 % women) with ages between 40 and 67 at baseline were followed until December 31st 2006. The data were linked with hospitalisation records and with the National Causes of Death Registry in order to use both incidence and mortality as endpoints. Baseline data included anthropometric measurements, blood pressure, serum lipids and questionnaires about lifestyle and health. The importance of the different risk factors was evaluated using Cox regression. In order to get hazard ratios (HR) on a common and comparable scale for all the risk factors, the latter were rank transformed and scaled to the range 0–5 before estimation. After transformation the HR can be interpreted as HR per quintile increment for each risk factor, and the relative sizes of the HRs can be used to rank the risk factors according to importance. We made separate models for four sex and age groups: Men =60, Women=60, Men>60 and Women >60.
Results: When cardiovascular death was considered as endpoint, the two most important risk factors for men = 60 were packyears of smoking and plasma homocysteine. For women <60 the same pattern was seen although the smoking effect was stronger than for men. In the younger age group serum total cholesterol and living alone were the most important risk factors for men while cholesterol and packyears were most important for women. The two most important risk factors for incident myocardial infarction (hospitalisations and death) for both male age groups were cholesterol and packyears and for women = 60 blood pressure and packyears. In the younger female age group the two most important risk factors were blood pressure and cholesterol.
Conclusion: We found different risk factor rankings for men and women and between the two age groups. Smoking was the most consistent risk factor for acute myocardial infarction incidence, followed by cholesterol; systolic blood pressure was significant only among women. In addition, living alone was a significant factor for cardiovascular death among men. Thus, rescaling risk factors to the same scale affords direct comparisons of their importance across gender and age groups.
P82
Asymmetrical distribution of atherosclerosis in the carotid artery: identical patterns across age, race and gender
P Tajik1, R Meijer1, GW Evans2, JJ Kastelein3, FJ Visseren4, JR Crouse Iii2, M Palmer1, ML Bots1
1University Medical Center Utrecht - Julius Centre for Health Sciences and Primary Care, Utrecht, Netherlands, 2Wake Forest University, Winston-Salem, United States of America, 3Academic Medical Center, Amsterdam, Netherlands, 4University Medical Center Utrecht, Department of Vascular Surgery, Utrecht, Netherlands
Topic: Cardiovascular epidemiology
Purpose: From autopsy studies among limited number of subjects and from clinical studies it has been indicated that carotid atherosclerosis develops in an asymmetrical manner. We set out to study circumferential distribution of carotid atherosclerosis and compare this pattern across various populations with different cardiovascular risk.
Methods: We used the baseline carotid intima-media thickness (CIMT) data from four recent international multicenter randomized controlled trials in which the carotid artery was systematically examined using the same ultrasound protocol and method to quantify CIMT. Of each subject circumferential information on the maximum CIMT of the left and right carotid arteries was obtained at common carotid, bifurcation and internal carotid segments. CIMT is presented in a circular manner and given by angle ranging from 0 to 360 degrees. At each segment, mixed modeling was used to study the differences in CIMT between angles, sides, gender, age, studies, ethnic and geographic origins.
Results: Each segment showed a different circumferential CIMT pattern. In all segments there was statistically significant difference between maximum CIMT across circumferential angles (p>0.001); on average CIMT was highest in posteromedial wall of bifurcation and internal carotid segment and at the anterolateral wall of the common carotid segment. The circumferential nonsymmetrical pattern was the same across side, gender, age, study, ethnic and geographic origin.
Conclusions: Apart from confirming asymmetrical atherosclerosis in the carotid artery, we expand the evidence by showing that the pattern of asymmetry is similar across populations with different vascular risks and across gender, age groups and races. Furthermore, our findings have serious implications for design of ultrasound protocols in CIMT studies.
P83
The danger of relying on univariate anthropometric measures for predicting major cardiovascular events
A Rosenblad, G Nilsson, J Leppert
Center for Clinical Research, Västerås, Sweden
Topic: Cardiovascular epidemiology
Purpose: To compare the univariate and multivariate predictive ability for Major Cardiovascular Events (MCVE) of the anthropometric measures Body Mass Index (BMI), Waist Circumference (WC), Waist-Hip Ratio (WHR), Waist-Height Ratio (WHtR) and Waist-Hip-Height Ratio (WHHR).
Methods: Men and women aged 40 and 50 from the county of Vastmanland, Sweden, were invited to a health survey during the years 1990–99 (participation rate 48%). MCVE (death or hospitalization with ICD-10:I21-I22, I62-I66 or I70-I74) was followed up until Dec. 2006. BMI, WC, WHR, WHtR and WHHR (defined as WHR divided by height) were calculated for persons with complete anthropometric measures (n=33493, 48% men). Cox regression was performed for each measure. Multivariate adjustment was performed for age, diabetic status, smoking, systolic and diastolic blood pressure, heart rate and glucose level. The univariate and multivariate hazard ratio (HR) P-values and predictive abilities of each measure were compared. The predictive ability was calculated with a Pseudo R-square measure. Multivariately the increase in Pseudo R-sq. when adding anthropometric measures was calculated.
Results: During follow up 1212 persons (3.6%, 815 men, 397 women) experienced MCVE. The univariate predictive ability was at most 22.6%, but this ability diminished when adjusting for confounders to at most 1.6%.
Conclusions: It is dangerous to rely on univariate methods for estimating the influence of anthropometric measures on MCVE for nonrandomised studies. We suggest that multivariate methods should be used for this kind of studies.
Univariate vs multivariate methods
Sex
Measure
Univariate HR P-value
Univariate Pseudo R-sq.
Multivariate HR P-value
Increase in multivariate Pseudo R-sq.∗
Men
BMI
>0.001
0.067
0.001
0.006
WC
>0.001
0.083
>0.001
0.007
WHR
>0.001
0.115
>0.001
0.010
WHtR
>0.001
0.116
>0.001
0.011
WHHR
>0.001
0.138
>0.001
0.013
Women
BMI
>0.001
0.122
0.033
0.004
WC
>0.001
0.194
0.001
0.011
WHR
>0.001
0.226
>0.001
0.016
WHtR
>0.001
0.208
>0.001
0.012
WHHR
>0.001
0.222
>0.001
0.014
∗Without anthropometric measures: Men 0.531, women 0.609.
P84
Association between C-reactive protein and alcohol consumption in Moscovites aged 55 and older
S A Shalnova1, AD Deev1, VA Metelskaya1, VM Shkolnikov2, MA Shkolnikova2, V Zhukova1
1National Research Center for Preventive Medicine, Moscow, Russian Federation, 2Max Planck Institute for Demographic Research, Rostock, Germany
Topic: Biomarkers
It is known that moderate alcohol consumption is associated with a lower carduiovascular risk. Elevated levels of C-reactive protein (CRP) as a marker of inflammation may explain this relationship.
Purpose: To investigate the relationship between CRP and alcohol consumption in Muscovites aged 55 and older.
Methods: The representative sample of 1,876 subjects (47.9% males and 52.1% females) took part in the baseline survey of the Stress, Aging and Health study in Russia (SAHR). The SAHR is a prospective population-based cohort study that is being conducted in Moscow. Alcohol intake was assessed by a standard frequency-amount questionnaire. Beer, wine, fortified wine and strong liqueurs were listed. CRP was measured by a highly sensitive particle-enhanced immunoturbidimetric assay (hsCRP). As a cut off point for high CRP, 3 mg/ml was taken (H3CRP). Socio-demographic characteristics, smoking status, anthropometry, blood pressure, lipid profile, glucose were included in the study protocol as well. Descriptive analysis included Mean±SD for continuous variables and relative frequencies for categorical variables. Logistic regression was applied for linking H3CRP with demographic, physiological and biochemical characteristics.
Results: Mean CRP in whole cohort was 3.26±.12, slightly higher in men then in women (3.38±.17 and 3.16±.16, respectively). CRP tended to increase with age, but not statistically significant. Abstainers constituted 20.9% of the sample (15.8% and 25.6% for males and females, respectively). About third of the subjects were H3CRP (34.1% and 32.2% for males and females, respectively). No statistically significant relationship between H3CRP and the total alcohol consumption was detected. But considering separately consumption of light beverages (LAC) such as beer, wine and strong liquors: vodka, cognac (SAC), we found negative (L-shaped) relation of H3CRP to LAC and still no relationship to SAC. Relative to abstainers, odds ratios [OR (95% CI)] linking H3CRP to LAC quartiles were: 0.801 (0.622,1.031); 0.536 (0.399, 0.720) (p>0.001); 0.667(0.501, 0.888) (p>0.001). After adjustment for age, sex, and conventional cardiovascular risk factors the respective OR estimates were: 0.786 (0.591, 1.073); 0.600(0.427, 0.841) (p>0.01); 0.709 (0.506, 0.996) (p>0.01).
Conclusion: Protective role of strong liquors in respect to inflammatory diseases is doubtful. There is, however, evidence linking consumption of beer and wine to lower levels of CRP. The fact that Russians prefer strong alcohols could reduces the protective role of moderate alcohol consumption.
