Demosthenes B. Panagiotakos a,b,, Christos Pitsavos b, Christina Chrysohoou b, John Skoumas b, Christodoulos Stefanadis b

Transcription

1 Atherosclerosis 173 (2004) Status and management of blood lipids in Greek adults and their relation to socio-demographic, lifestyle and dietary factors: the ATTICA Study Blood lipids distribution in Greece Demosthenes B. Panagiotakos a,b,, Christos Pitsavos b, Christina Chrysohoou b, John Skoumas b, Christodoulos Stefanadis b a Department of Dietetics and Nutrition, Harokopio University, 46 Paleon Polemiston St. Glyfada, Athens , Greece b First Cardiology Clinic, School of Medicine, University of Athens, Athens, Greece Received 11 September 2003; accepted 23 December 2003 Abstract Objective: In this work, we assessed the status and management of blood lipids in a sample of cardiovascular disease free adult men and women from Greece. We also evaluated the effect of several socio-demographic, dietary and lifestyle habits on lipid levels. Methods: The ATTICA Study is a population-based cohort that has randomly enrolled 1128 men and 1154 women (aged >18 years old), stratified by age gender, from the greater area of Athens, during Adherence to Mediterranean diet was assessed through a diet score that was based on a validated food-frequency questionnaire. Results: Forty-six percent of men and 40% of women had total serum cholesterol levels >200 mg/dl. Of them, 40% of men and 30% of women were unaware of their condition. Twenty-one percent of men and 7% of women had HDL cholesterol levels <35 mg/dl. Twenty-eight percent of men and 13% of women had triglyceride levels >150 mg/dl. Fifteen percent of men and 12% of women had LDL cholesterol levels >160 mg/dl and 52% of men and 48% of women had LDL >130 mg/dl. Of those who had known blood lipid abnormalities, 36% of men and 33% of women followed a dietary medication, 31% of men and 20% of women were receiving a pharmaceutical treatment (mainly statin) and the rest were untreated. Participants who adopted the Mediterranean diet and received statin, had on average 9% lower total cholesterol (P = 0.04), 19% lower LDL cholesterol levels (P = 0.02) and 32% lower oxidized LDL cholesterol levels (P <0.001) compared to those who were untreated and adopted a Westernized diet. Conclusions: We could speculate that about 3 million Greek adults had high total cholesterol levels. Adverse findings were also observed regarding the other investigated blood lipids. Mediterranean diet could be a complimentary mean to pharmaceutical treatment in reducing blood lipids Elsevier Ireland Ltd. All rights reserved. Keywords: Blood lipids; Mediterranean diet; Atherosclerosis 1. Introduction During the past years much scientific evidence has been raised regarding the adverse effect of abnormal blood lipid levels on atherosclerotic disease. At first, both epidemiological and clinical studies revealed a relation between total cholesterol and the risk of coronary heart disease, due to its feasibility in field investigations [1]. However, there is now a substantial body of information on distributions and the predictive importance for coronary artery disease of other blood lipids and lipoproteins [1,4]. At this point it should be Corresponding author. Tel.: ; fax: address: d.b.panagiotakos@usa.net (D.B. Panagiotakos). noted that the strength of the relationships between blood lipids and atherosclerosis might be influenced by several other lifestyle-related factors. The Framingham Heart Study first reported that glucose intolerance, blood pressure levels and smoking habits could influence the effect of cholesterol on coronary risk [1]. In addition, socio-economic studies [5] suggest an inverse association between social status and the distribution of various blood lipids. It is also well known that dietary patterns are strongly related to the prevalence and the management of dyslipidemia [1,6,7]. For centuries Greeks followed to consume a traditional dietary pattern, the Mediterranean diet [8,9]. This dietary pattern has received special attention during the past years [10 12]; however, its relation with the prevalence and management of dyslipidemia is not well understood /$ see front matter 2004 Elsevier Ireland Ltd. All rights reserved. doi: /j.atherosclerosis

2 354 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) Given the lack of current data regarding the levels of blood lipids in Greek population, we investigated the distribution, awareness and management of total, HDL, LDL, oxidized LDL cholesterol, triglycerides and lipoprotein(a) (Lp(a)) levels, in a random sample of cardiovascular disease free adults. Moreover, we evaluated the association between adherence to the Mediterranean diet and blood lipid levels, after controlling for several socio-demographic and lifestyle factors. 2. Methods 2.1. Population of the study The ATTICA Study is a health and nutrition survey, which is being carried out in the province of ATTICA (including 78% urban and 22% rural areas), where Athens is a major metropolis. The sampling was random, multistage and based on the age sex distribution of the province of AT- TICA provided by the National Statistical Service (census of 2001). The sampling anticipated enrolling only one participant per household. Also, all people living in institutions were excluded from the study. From May 2001 to August 2002, 3355 inhabitants from the above area were randomly selected to enter the study. Of them, 2282 agreed to participate (68% participation rate). The selected sample can be considered as representative since there were only minor, insignificant, differences in sex and age distribution between the study population and the target population. All participants were interviewed by trained personnel (cardiologists, dieticians and nurses) who used a standard questionnaire. The number of enrolled participants is adequate to evaluate differences between the investigated parameters greater than 20%, achieving statistical power >0.80 at <0.05 probability level (P-value) Dietary assessment Consumption of non-refined cereals and products, vegetables, legumes, fruits, olive oil, dairy products, fish, pulses, nuts, potatoes, eggs, sweets, poultry, red meat and meat products were measured as an average per week during the past year through a validated food-frequency questionnaire, from the Department of Nutritional Epidemiology of our Institute [13]. The frequency of consumption was then quantified approximately in terms of the number of times per month a food was consumed. Alcohol consumption was measured by daily ethanol intake, in wineglasses (100 ml and 12% ethanol concentration). The Mediterranean dietary pattern consists of: (a) daily consumption of non-refined cereals and products (whole grain bread, pasta, brown rice, etc.), vegetables (2 3 servings per day), fruits (6 servings per day), olive oil (as the main added lipid) and dairy products (1 2 servings per day), (b) weekly consumption of fish (4 5 servings per week), poultry (3 4 servings per week), olives, pulses, and nuts (3 servings per week), potatoes, eggs and sweets (3 4 servings per week) and monthly consumption of red meat and meat products (4 5 servings per month). It is, also, characterized by moderate consumption of wine (1 2 wineglasses per day) and high monounsaturated: saturated fat ratio (>2). A Harvard-led group suggested this dietary pyramid with substantial input from Greek scientists [14]. Based on this dietary pyramid, we calculated a special diet score that assessed adherence to Mediterranean diet. In particular, for the consumption of food items that are close to this dietary pattern we assigned score 0 for rare or no consumption, 1 for 1 4 times per month, 2 for 5 8 times, 3 for 9 12 times per month, 4 for times per month and 5 for almost daily consumption. On the other hand, for the consumption of foods that are away from this traditional diet, like meat and meat products, we assigned the opposite scores (i.e. 0 for almost daily consumption to 5 for rare or no consumption). For alcohol, we assigned score 5 for consumption of less than three wineglasses per day and, progressively, score 0 for consumption of more than seven wineglasses per day. Higher values of the suggested dietary score indicates adherence to the traditional Mediterranean diet (i.e. which is also characterized by moderate consumption of fat and high monounsaturated: saturated fat ratio), while lower values indicate adherence to the Westernised diet. We defined subjects who are closer to the Mediterranean type of diet using as cut-off points the median values of this food consumption score as done by several investigators [12,14] Biochemical measurements The blood samples were collected from the antecubital vein between 8 a.m. and 10 a.m., in a sitting position, after 12 h of fasting and avoiding alcohol. The biochemical evaluation was carried out in the same laboratory that followed the criteria of the World Health Organization Lipid Reference Laboratories. All biochemical examinations (serum total cholesterol, HDL cholesterol and triglycerides) were measured using chromatographic enzymic method in a Technicon automatic analyser RA-1000 (Dade Behring, Marburg, Germany). Oxidized LDL cholesterol was measured in plasma using enzyme-linked immunosorbent assay (ELISA) kit (Mercodia AB, Uppsala, Sweden). HDL cholesterol was determined after precipitation of the Apolipoprotein B containing lipoproteins with dextran-magnesium-chloride. Non-HDL cholesterol was calculated by the formula: total cholesterol minus HDL cholesterol. Lipoprotein(a) was measured by a latex enhanced turbidimetric immunoassay. Serum for the measurement of these lipids was harvested immediately after admission. LDL cholesterol calculated using the Friedewald formulae: {total cholesterol} {HDL cholesterol} (1/5)(triglycerides). Non-HDL cholesterol is the total cholesterol minus HDL cholesterol. An internal quality control was in place for assessing the validity of cholesterol, triglyceride and HDL methods. The

3 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) intra and inter-assay coefficients of variation of cholesterol levels did not exceed 9%, triglycerides 4% and HDL 4% Demographic, clinical and lifestyle characteristics The study s questionnaire also included demographic characteristics like age, gender, financial status (average annual income during the past 3 years), and education level. The educational level of the participants (as a proxy of social status) was measured by the years of school (Group I: <9 years, Group II: up to high school or technical colleges (10 14 years) and Group III: university). Information about smoking habits was collected using a standardized questionnaire developed for the study. Current smokers were defined as those who smoked at least one cigarette per day, never smokers, those who have never tried a cigarette in their life; and former smokers were defined as those who had stopped smoking more than 1 year previously. For the multivariate statistical analyses cigarette smoking was quantified in pack years (cigarette packs per day X years of smoking), adjusted for a nicotine content of 0.8 mg per cigarette. Physical activity was defined as leisure-time activity of a certain intensity and duration, at least once per week during the past year, and was graded in qualitative terms such as light (expended calories <4 kcal/min), moderate (expended calories 4 7 kcal/min) and vigorous (expended calories >7 kcal/min). The rest of the subjects were defined as physically inactive. Body mass index was calculated as weight (in kg) divided by standing height (in meters squared). Obesity was defined as body mass index >29.9 kg/m 2. Arterial blood pressure was measured three times at the right arm (ELKA aneroid manometric sphygmometer, Von Schlieben Co, West Germany), at the end of the physical examination with subject in sitting position at least for 30 min. The systolic blood pressure level was determined by the first perception of sound (of tapping quality). The diastolic blood pressure level was determined by phase V when the repetitive sounds become fully muffled (disappear). Changes in loudness were not considered. Patients whose average blood pressure levels were greater or equal to 140/90 mmhg or were receiving antihypertensive medication were classified as hypertensives. Hypercholesterolemia was defined as total serum cholesterol levels greater than 220 mg/dl or the use of lipid-lowering agents and diabetes mellitus as a blood sugar >125 mg/dl or the use of antidiabetic medication. Details regarding the aims and the design of the ATTICA Study have been presented already, elsewhere [11,15] Statistical analysis Continuous variables are presented as mean values ± 1 standard deviation, while qualitative variables are presented as absolute and relative frequencies. Associations between categorical variables were tested by the use of contingency tables and the calculation of Chi-squared test. Comparisons between normally distributed continuous and categorical variables were performed by the calculation of Student s t-test and one-way or multi-way analysis of co-variance (MANCOVA), after testing for equality of variances (homoscedacity). In the case of asymmetric continuous variables the tested hypotheses were based on the calculations of the non-parametric test suggested by Kruskal and Wallis. Correlations between lipid levels and age, smoking habits (in pack years) and body mass index were evaluated by the calculation of Pearson s correlation coefficient for the normally distributed variables and by the Spearman correlation coefficient for the skewed variables. Differences in blood lipids between people who were close to versus those who were away from the Mediterranean diet were tested using MANCOVA, after taking into account the effect of several potential confounders. Furthermore, the previous associations were also tested through multiple linear regression analysis after the adjustment for age, sex, body mass index, smoking habits, physical activity status, medication and education status and their interactions. Introducing alternately the variables representing the food groups into the core model that included the other investigated characteristics of the participants assessed the effect of qualitative aspects of nutrition on blood lipids. All reported P-values are based on two-sided tests and compared to a significance level of 5%. SPSS 11.0 Software (SPSS Inc., 2002, IL, USA) was used for all the statistical calculations. 3. Results Demographic, clinical and behavioral characteristics of the participants are presented in Table Blood lipids distribution The mean values of the investigated blood lipids as well as the percentage of abnormal levels both in men and women are presented in Table 2. In addition to this information, 15% of men and 13% of women had cholesterol levels above 240 mg/dl. As expected, when we focused our analysis in older adults (>50 years old) we observed that 48% of men and 55% of women had total cholesterol levels >200 mg/dl (results not shown in Table 2). Similar results were observed regarding HDL cholesterol (Table 2). We also observed that 21% of men and 7% of women had very low HDL cholesterol levels (<35 mg/dl). A group of people of particular interest is those who have normal total cholesterol, but low HDL cholesterol levels. In our population, 19% of men and 12% of women who had normal total cholesterol levels (i.e. <200 mg/dl) had HDL cholesterol levels lower than 35 and 45 mg/dl, respectively. Finally, 52% of men and 48% of women had LDL cholesterol levels greater than 130 mg/dl (results not shown in Table 2). It is known that age is a factor that correlates well with blood lipid levels. In our study, age was positively and significantly associated with all blood lipids measurements in both

4 356 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) Table 1 Characteristics of the study s participants (% by gender) Men (1128% or 49%) Women (1154% or 51%) Age (years) 47 ± ± Education status Group I 245 (22%) 290 (25%) Group II 429 (38%) 440 (38%) Group III 454 (40%) 424 (37%) Smoking habits <0.001 Never smoker 357 (32%) 638 (55%) Current smoker 544 (48%) 441 (38%) Former smoker 227 (20%) 75 (7%) Physical activity status Sedentary 664 (59%) 722 (63%) Light or moderate 234 (21%) 177 (15%) Vigorous 230 (20%) 255 (22%) Hypertension 431 (38%) 276 (24%) <0.001 Diabetes mellitus 107 (10%) 82 (7%) Obesity 272 (25%) 194 (17%) <0.001 Diet score (range 0 55) 35 ± ± 12 <0.001 Alcohol consumption (ml per day) 220 ± ± 40 <0.001 men and women, with the exception of HDL cholesterol levels (Fig. 1). Thus in this paragraph we expanded our findings by presenting the distribution of the investigated blood lipids after adjusting for age, and controlling for other potential confounders that correlate with age, like sex, smoking habits, physical activity status and diet score (results not shown in tables or figure). We observed that a decade difference in age was associated with 9 mg/dl higher total cholesterol levels (95% confidence interval (CI) from 8 to 11 mg/dl, P< 0.001), 13 mg/dl higher triglycerides levels (95% CI from 11 to 16 mg/dl, P<0.001), 7 mg/dl higher LDL cholesterol P levels (95% CI from 5 to 9 mg/dl, P<0.001), 1.2 mg/dl higher lipoprotein-(a) levels (95% CI from 0.2 to 1.9 mg/dl, P<0.001), while only 0.02 lower HDL cholesterol levels (95% CI from 0.1 to 0.2 mg/dl, P = 0.352). Since several investigators claimed an effect of social status on blood lipid levels we evaluated the distribution of lipids in relation to education status as a proxy of socio-economic level of the participants. After controlling for several potential confounders, like age, sex, body mass index, physical activity levels, smoking and dietary habits, we found that only HDL cholesterol levels had a consistent positive association with education status (data not presented in tables). In particular, compared to lower educated (Group I) people in the highest education group (Group III) had on average 2 mg/dl higher HDL cholesterol values (95% CI: mg/dl, P = 0.04). No statistically significant associations were observed between blood lipids levels and income of the participants. Finally, we aimed to evaluate the distribution of blood lipids in regard to physical activity status and body mass index of the participants. However, with the exception of HDL levels in women (where physically active subjects had on average 4 mg/dl higher levels as compared to sedentary, P = 0.024), no other significant associations were observed between blood lipids and physical activity status, after controlling for age and body mass index of the participants. In addition, compared to normal or overweight obese participants had on average 25 mg/dl higher triglycerides (P <0.001) and 4 mg/dl lower HDL levels (P <0.001), after controlling for age, sex, physical activity status, dietary and smoking habits. No other significant associations as regards blood lipid levels and body mass index of the participants were observed, after controlling for the aforementioned confounders. Table 2 Blood lipids distribution in men and women without any clinical evidence for cardiovascular disease Men (1128% or 49%) Women (1154% or 51%) P Mean ± S.D. 5% 95% Mean ± S.D. 5% 95% Percentile Total cholesterol (mg/dl) 199 ± ± % of people with total cholesterol >200 mg/dl Triglycerides (mg/dl) 138 ± ± <0.001 % of people with triglycerides >150 mg/dl HDL cholesterol (mg/dl) 44 ± ± <0.001 % of men/women with HDL <40/50 mg/dl Non-HDL cholesterol (mg/dl) 156 ± ± <0.001 % of men/women with non-hdl >190 mg/dl LDL cholesterol (mg/dl) 128 ± ± % of men/women with LDL >160 mg/dl Oxidized LDL cholesterol (mg/dl) 68 ± ± Lipoprotein(a) (mg/dl) 18 ± ± % of people with Lp(a) >30 mg/dl P-value derived from the comparison between men vs. women after taking into account the effect of age, body mass index, physical activity status and smoking habits.

5 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) Fig. 1. Selected blood lipids and age of participants (lines represent regression line and 95% confidence interval) Blood lipids management Of those who had already known blood lipid abnormalities (i.e. 240 men and 280 women), 36% of men and 33% of women followed only dietary treatment and 31% of men and 20% of women were under a complimentary pharmaceutical medication. Thus, 33% of men and 47% of women with known lipid abnormalities were untreated. Moreover, of those who were found to have abnormal total or LDL cholesterol levels at interview, 40% of men and 30% women reported that they were unaware about their condition and they had never had their blood lipid levels measured before. Moreover, people with known but untreated lipid abnormalities had lower education status as compared to those who were receiving dietary or pharmacological treatment (10±4 versus 12±4 years of school, P<0.001). Statin was the main lipid-lowering agent (55% received atorvastatin, 15% pravastatin, 12% simvastatin and 10% another type of statin), followed by fibrates (7%) and other lipid-lowering drugs ( 1%). However, only the 49% of men and the 51% of women who were under pharmacological treatment had desirable total cholesterol levels (i.e. <200 mg/dl) at interview Blood lipids and dietary characteristics A secondary goal of this work was to assess blood lipid levels with dietary habits. We initially evaluated the effect of Mediterranean diet on the investigated lipids (Table 3). We observed that participants who adopted the Mediterranean diet had similar blood lipid profile compared to those who were closer to a Westernised diet, after controlling for

6 358 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) Table 3 Blood lipid levels in relation to the adherence of the Mediterranean diet Mediterranean diet Westernized diet P Men Women Men Women Total cholesterol (mg/dl) 200 ± ± ± ± Triglycerides (mg/dl) 143 ± ± ± ± HDL cholesterol (mg/dl) 43 ± ± ± ± LDL cholesterol (mg/dl) 130 ± ± ± ± Oxidized LDL cholesterol (mg/dl) 65 ± ± ± ± Lipoprotein(a) (mg/dl) 19 ± ± ± ± P-values derived from the comparison between people who were close to the Mediterranean diet vs. those who were closer to a Westernized diet, after taking into account the effect of gender, age, body mass index, smoking habits, education and physical activity status. several potential confounders presented in Table 3. Only the differences observed in oxidized LDL cholesterol levels reached statistical significance (Table 3). Furthermore, multivariate analysis showed that a 10-unit increase in diet score was associated with 22 mg/dl (95% CI: 8 36, P = 0.04) lower oxidized LDL cholesterol levels, after controlling for age, sex, body mass index, smoking habits, education and physical activity status of the participants (data not presented in Table 3). Additional food group analysis (not presented in Table 3) showed that compared to no consumption those in the highest quartile of legumes consumption (i.e. >3 times per week) had 19% lower LDL cholesterol levels (Fig. 2, P = 0.04) and 6% higher HDL cholesterol levels (P = 0.03). Furthermore, fruit and vegetable intake was associated with 13% lower total cholesterol levels (P = 0.04), 20% lower triglycerides (P = 0.02) and 8% higher HDL cholesterol levels (P = 0.03). Afterwards we focused our interest on participants who were under lipid-lowering treatment and also adopted the Mediterranean diet (data not presented in Table 3). We found that the aforementioned combination was associated with 9% reduction in total serum cholesterol levels (P = 0.04). In addition, participants who adopted the Mediterranean diet and received statin had on average 19% lower LDL cholesterol levels (P = 0.021) and 32% lower oxidized LDL cholesterol levels (P < 0.001) compared to those who were untreated and closer to a Westernised diet. Of particular interest is that the observed reductions were higher than the reductions achieved by diet or statin treatment alone (P for interaction <0.05). The effect of this combination on the other rest lipids was not statistically significant. 4. Discussion In this work, we evaluated the distribution of blood lipid in a random, population-based, sample of Greek adults. We observed that a considerable proportion of the participants had high blood lipid levels, increasing substantially in older adults. Moreover, roughly four out of ten participants with abnormal blood lipids were untreated, and only half of those who used lipid-lowering agents reached optimal lipid levels. Although the latter finding may be influenced by the compliance to medication and other confounders, it is of great importance for the public health strategies in the investigated population. Regarding the various dietary characteristics, we observed that adherence to the Mediterranean diet resulted in a significant reduction only in oxidized LDL cholesterol levels. In addition, we revealed the synergistic effect of Mediterranean diet with statin treatment in the management of blood lipids. Fig. 2. LDL cholesterol levels and legumes consumption.

7 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) Epidemiology of blood lipids Total cholesterol: Forty-six percent of men and 40% of women had total cholesterol levels greater than 200 mg/dl. Based on these data, it could be speculated that 1.6 million adult men and 1.4 million adult women in Greece have high total cholesterol levels. Our findings were slightly different from those reported by the National Health and Nutrition Examination Survey (NHANES) III study [16], where 52% of non-hispanic white adult men and 49% of women in USA had total blood cholesterol levels over 200 mg/dl, and 18% of men and 20% of women had total blood cholesterol levels of 240 mg/dl or higher. Studies [1 3] show that a higher percentage of women than men have total blood cholesterol of 200 mg/dl or higher, beginning at age 50. The latter was confirmed by our study. Current data regarding the total cholesterol levels in Greek population are lacking. The investigators from the Seven Countries Study [17] reported that in the early 1980s more than 50% of Athenian men and women, between 40 and 59 years old, had total cholesterol levels greater than 240 mg/dl. In another study [18], also conducted in Athens during the middle 1980s, the investigators reported that hypercholesterolemia (total cholesterol >240 mg/dl) prevailed in 20% of men and 17% of women adults. Thus, there is evidence to support that the proportion of abnormal total cholesterol levels increased during the past two decades in our population. This may attribute to the adherence of a more Westernised lifestyle that observed in Greece during the past two decades, which includes consumption of high-fat foods, adoption of sedentary life and increased cigarette smoking [17]. HDL cholesterol: A wide range of scientific evidence suggests that HDL cholesterol plays a role in the development of coronary heart disease [1,3,19]. The NHANES III study [16] reported that 18% of men and 6% of women had HDL cholesterol levels below 35 mg/dl, and similar results were found in our study. Studies suggest that even for those with normal levels of total cholesterol, risk for myocardial infarction is high when HDL cholesterol is low. We observed that 19% of men and 12% of women, who had desirable total cholesterol levels, had HDL cholesterol levels lower than 35 and 45 mg/dl, respectively. The latter may underline the importance of total-to-hdl cholesterol ratio for the evaluation of blood lipids and the prevention of atherosclerotic disease, at population level. LDL cholesterol: According to several observational and clinical studies LDL cholesterol levels of 160 mg/dl or greater have a significant contribution for the development and the progression of coronary heart disease [1]. We observed that 15% of men and 12% of women had LDL cholesterol levels above 160 mg/dl. In the NHANES III [16], these rates were 22 and 17%, respectively. Moreover, a considerable proportion of men and women had LDL cholesterol levels greater than 130 mg/dl. Recently, the Adult Treatment Plan (ATP) III [4] based initiation and treatment goals for dietary and pharmacological therapy on LDL cholesterol levels, number of pre-existing risk factors and previous experience of coronary heart disease. According to these guidelines, individuals without coronary heart disease and fewer than two risk factors should initiate dietary therapy when LDL cholesterol levels exceed 160 mg/dl. Taking into account the high prevalence of the cardiovascular risk factors in our population, it seems that a considerable proportion of men and women should be under lipid-lowering agents. Triglycerides: The ATP III suggests a cut off point of 150 mg/dl for defining elevated triglycerides levels. Few years ago a large meta-analysis [20] showed that increase in triglyceride levels was independently associated with increased coronary risk in men and women. In our study, approximately one third of men and only 13% of women had triglycerides levels of 150 mg/dl or higher. Thus, approximately 980,000 of adult men and 455,000 of women had high triglycerides levels. Unfortunately, population based data from other relevant studies regarding triglycerides levels are lacking; this makes the comparisons with our findings difficult. Nevertheless, it is noteworthy that a considerable proportion of men and women are at risk because of their triglyceride levels. Lipoprotein(a): It has been reported that the distribution of lipoprotein(a) levels in the population are different from the bell-shaped curve of serum cholesterol [21]. Thus, the majority of people in the US have low levels (less than 10 mg/dl), while increased prevalence of high lipoprotein(a) levels, i.e. greater than 30 mg/dl, are more common in the Asian Indian and Turkish populations [22]. The mean lipoprotein(a) values in our sample were between the US and east Mediterranean and Asian populations (Table 2). Moreover, approximately 15% of men and women had lipoprotein(a) levels higher than 30 mg/dl. The most recent population-based data on lipoprotein(a) levels in Greece have been reported by Dionyssiou-Asteriou and Rizos [23]. In particular, studying a small sample of Greek adults they observed that 23% of men and 34% of women had lipoprotein(a) level above 30 mg/dl. In the present work studying a much larger population sample of men and women we observed much lower rates. These differences may attribute to the different sampling procedures, the size of the investigated samples, or the laboratory techniques used. Moreover, we observed that women had significantly higher values than men, which was in accordance to the gender differences observed by the previous investigators [23] Dietary and pharmacological management of blood lipids It is known that dietary habits usually influence blood lipids [7]. Thus, we evaluated lipid levels under the prism of the adoption or not of the traditional Mediterranean diet, as well as its constituents. This diet has already been related with the reduction of all cause and cardiovascular disease mortality, due to its effect on blood pressure levels,

8 360 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) body mass index, platelet aggregation, plasma fibrinogen and other haemostaseological factors [8 11]. Moreover, Petridou et al. [24] studying blood samples from Greek adolescents reported that the traditional Mediterranean pattern of living and eating was associated with a favorable lipid profile. However, benefits from this dietary pattern on blood lipids have rarely been reported in the literature. In our study, we showed that only on oxidized LDL cholesterol levels, adoption of Mediterranean diet resulted a significant reduction. In particular, people who adopted this diet had on average lower levels as compared to those who adopted a Westernized type of diet, after controlling for several potential confounders. Thus, based on this observation we could state another hypothesis of a pathophysiological mechanism by which Mediterranean diet may reduce cardiovascular risk, through the moderation of the oxidation process. Moreover, we revealed the additive effect of Mediterranean diet with statin treatment on blood lipids. Since the levels of uncontrolled or untreated dyslipidemia seem high in our population the previous finding could be of high public health interest. However, the latter could be confounded by the better compliance to the treatment by people who were closer to the Mediterranean dietary pattern; which could not be assessed by the present study. In addition, increased consumption of legumes, fruits and vegetables was associated with significant reduction on several blood lipid levels. These findings may suggest another non-pharmacological intervention that may reduce blood lipid levels. However, several methodologic reasons underline the inherent difficulty to demonstrate associations between specific dietary intakes and serum lipids, mainly because endogenous variations overwhelm exogenous components, which are measured in sub-optimal way [24]. Thus, the associations observed in the present study should be considered in a more thoughtful way. At this point it should be noted that the extrapolation of our findings into the general population might be under scrutiny. One of the main reasons is the moderate participation rate (68%), which is acceptable for population-based studies, like the present one, but may state hypotheses that the lifestyles of those who agreed to participate and of those who did not could be different. One other limitation is the cross-sectional design of the study. Thus, the observed benefits from the Mediterranean diet on oxidized LDL cholesterol levels, or the effect of physical activity on HDL cholesterol levels should be further investigated by randomized clinical trials. 5. Conclusion Despite the aforementioned limitations, the present study revealed that in the 21st century a considerable proportion of men and women have abnormal blood lipid levels. Based on our findings, we can speculate that roughly 3.0 million Greeks, have high total cholesterol levels. In addition, more than one out of three participants who had abnormal lipid levels were unaware of their condition, while only one out of two who receiving lipid-lowering agents reached desirable blood lipid levels. As regards the effect of dietary habits, although adherence to a Mediterranean diet did not influence blood lipids (with the exception of oxidized LDL cholesterol) the combination with statin treatment showed an additive favorable effect on the investigated lipids. Based on the likelihood that modification of lipid levels will be beneficial, especially, to those at higher risk for coronary heart disease, we suggest that screening for these abnormalities is considered essential and must be followed by active and effective interventions [25]. Acknowledgements The ATTICA Study is supported by research grants from the Hellenic Cardiological Society (HCS2002) and the Hellenic Atherosclerosis Society (HAS2003). The authors would like to thank the field investigators of ATTICA Study: Dr. Natasa Katinioti (physical examination), Dr. Akis Zeimbekis (physical examination), Dr. Spiros Vellas (physical examination), Dr. Efi Tsetsekou (physical/psychological evaluation), Dr. Dina Masoura (physical examination), Dr. Lambros Papadimitriou (physical examination), as well as the technical team: Dr. Marina Toutouza (senior investigator/biochemical analysis), Ms. Carmen Vasiliadou (genetic analysis), Mr. Manolis Kambaxis (nutritional evaluation), Ms. Konstadina Paliou (nutritional evaluation), Ms. Constadina Tselika (biochemical evaluation), Ms. Sia Poulopoulou (biochemical evaluation) and Ms. Maria Toutouza (database management). References [1] Gotto Jr AM, Lipid and lipoprotein disorders. In: Pearson TA, Criqui MH, Luepker RV, Oberman A, Wilson M, editors. Primer in preventive cardiology. Dallas, Tex: American Heart Association;1994. p [2] Ginsberg HN. Lipoprotein metabolism and its relationship to atherosclerosis. Med Clin North Am 1994;78:1 20. [3] Wilson PWF, Abbott RD, Castelli WP. High density lipoprotein cholesterol and mortality. Arteriosclerosis 1988;8: [4] Executive summary of the third report of the National Cholesterol Educational Program Expert panel on detection, evaluation, and treatment of high blood cholesterol in adults. JAMA 2001; 285(19): [5] Gran B. Major differences in cardiovascular risk indicators by educational status. Results from a population based screening program. Scand J Soc Med 1995;23:9 16. [6] Keys A, Menotti A, Karvonen MJ. The diet and 15-year death rate in the Seven Countries Study. Am J Epidemiol 1986;124: [7] Denke M. Cholesterol lowering diets. A review of evidence. Arch Int Med 1995;155: [8] Trichopoulou A, Kouris-Blazos A, Wahlqvist M, et al. Diet and overall survival in elderly people. Brit Med J 1995;311: [9] Kafatos A, Diacatou A, Voukiklaris G. Heart disease risk-factor status and dietary changes in the Cretan population over the past

9 D.B. Panagiotakos et al. / Atherosclerosis 173 (2004) years: the Seven Countries Study. Am J Clin Nutr 1997;65: [10] de Lorgeril M, Salen P, Martin J-L, Monjaud I, Delaye J, Mamelle N. Mediterranean diet, traditional risk factors and the rate of cardiovascular complications after myocardial infarction. Final report of the Lyon Diet Heart Study. Circulation 1999;99: [11] Panagiotakos DB, Pitsavos H, Chrysohoou C, et al. Status and management of hypertension, in Greece the role of the adoption of Mediterranean diet: the ATTICA Study. J Hypertens 2003;21: [12] Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean diet and survival in a Greek population. N Engl J Med 2003;348: [13] Trichopoulou A. From research to education: the Greek experience. Nutrition 2000;16: [14] Willett WC, Sacks F, Trichopoulou A, et al. Mediterranean diet pyramid: a cultural model for healthy eating. Am J Clin Nutr 1995;6:1402S 6S. [15] Pitsavos C, Chrysohoou C, Panagiotakos DB, et al. Association of leisure-time physical activity on inflammation markers (C-reactive protein, white-blood cell count, serum amyloid A and fibrinogen), in healthy subjects (from the ATTICA Study). Am J Cardiol 2003;91: [16] Third report on nutrition monitoring in the United States, vol. I. Washington, DC: Government Printing Office; [17] Dontas A. Recent trends in cardiovascular disease and risk factors in the Seven Countries Study: Greece. In: Toshima H, Koga Y, Blackburn H, Keys A, editors. Lessons for science from the Seven Countries Study. Tokyo, Japan: Springer-Verlag; [18] Moulopoulos SD, Adamopoulos PN, Diamantopoulos EI, Nanas SN, Anthopoulos LN, Iliadi-Alexandrou M. Coronary heart disease risk factors in a random sample of Athenian adults. The Athens Study. Am J Epidemiol 1987;126: [19] Franceschini G. Epidemiologic evidence for high-density lipoprotein cholesterol as a risk factor for coronary artery disease. Am J Cardiol 2001;88:9N 13N. [20] Hokanson JE, Austin MA. Plasma triglyceride levels is a risk factor for cardiovascular disease independent of high-density lipoprotein cholesterol level: a meta-analysis of population based prospective studies. J Cardiovasc Risk 1996;3: [21] Bostom AG, Cupples LA, Jenner JL. Lipoprotein(a) levels and risk for coronary heart disease in men aged 55 years old and younger a prospective study. JAMA 1996;276: [22] Davidson M. The mobile lipid clinic. Philadelphia, USA: Lippincot Williams and Wilkins Publications; [23] Dionyssiou-Asteriou A, Rizos I. Serum lipoprotein-(a) levels in a Greek population sample without a history of premature myocardial infarction. J Cardiovasc Risk 1996;3: [24] Petridou E, Malamou H, Doxiadis S, et al. Blood lipids in Greek adolescents and their relation to diet, obesity, and socioeconomic factors. Ann Epidemiol 1995;5: [25] Shepherd J. Economics of lipid lowering in primary prevention: lessons from the West of Scotland Coronary Prevention Study. Am J Cardiol 2001;87:19 22.