Introduction Materials and methods

Atherosclerosis 153 2000 107 – 117 Catechin in the Mediterranean diet: vegetable, fruit or wine? Jean-Bernard Ruidavets a, , Pierre-Louis Teissedre b , Jean Ferrie`res a , Ste´phane Carando b , Guillaume Bougard c , Jean-Claude Cabanis b a INSERM U 518 , De´partement d ’ e´pide´miologie, Faculte´ de me´decine, 37 Alle´es Jules Guesde, 31073 Toulouse Cedex, France b Laboratoire de chimie analytique, Faculte´ de Pharmacie, Uni6ersite´ de Montpellier, Montpellier, France c Ser6ice commun de spectrome´trie de masse, Faculte´ de Pharmacie, Uni6ersite´ de Montpellier, Montpellier, France Received 7 June 1999; received in revised form 12 November 1999; accepted 16 December 1999 Abstract The aim of this study was to determine which type of diet contributes most to plasma concentration of + -catechin, a naturally occurring antioxidant flavonoid. Consecutive subjects n = 180 were screened. A blood sample was collected after a fasting period and + -catechin measurement in plasma was performed by high-performance liquid chromatography HPLC method using fluorescence detection. Dietary consumption of the last evening meal was assessed by a dietary recall method. Taking fruit, vegetable and wine consumption into account, four types of diet were identified. After adjustment for confounding factors, concentration of + -catechin in plasma was three-fold higher in diet with fruit and vegetable but without wine 449.5 mgl, and four-fold higher in diet with wine but without vegetable and fruit 598.5 mgl in comparison to diet without fruit, vegetable and wine 131.6 mgl. When the consumption of vegetable, fruit and wine was combined, the concentration was the highest 637.1 mgl P B 0.001. Vegetable, fruit and wine were the major determinants of plasma + -catechin concentration P B 0.001. This study demonstrates that the highest plasma concentration of + -catechin was observed in subjects consuming fruit, vegetable and wine, and its antioxidant and antiaggregant activity could partly explain the relative protection against coronary heart disease CHD. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords : Catechin; Flavonoid; Antioxidant; Diet; Wine www.elsevier.comlocateatherosclerosis

1. Introduction

Flavonoids from vegetable and fruit intake appear to be inversely related to coronary heart disease CHD mortality [1 – 4]. Flavonoids from red wine have been shown to strongly inhibit low-density lipoprotein oxida- tion [5 – 8] both in vitro and in vivo, but also to reduce platelet aggregation [9,10], i.e. significant steps in reduc- ing CHD mortality. In the same way, + -catechin, a naturally occurring flavonoid, has been demonstrated to prevent human plasma oxidation [11] and to inhibit oxidation of low-density lipoprotein [12]. Thus, flavonoids may partly explain the protective effects of the Mediterranean diet [13], rich in vegetable, fruit and wine [14], against CHD. The first step in evaluating this hypothesis is to determine to what extent each of these foodstuffs contributes to antioxidant flavonoid levels in the blood. The aim of this study was to determine which type of diet contributes most to plasma concen- tration of + -catechin.

2. Materials and methods

2 . 1 . Population and sampling A cross-sectional study was carried out from June 1996 to April 1997. A sample of 1183 men and women aged 35 – 65 years and living in the region of Toulouse south-west of France, an area with low CHD mortal- ity and morbidity rates, was recruited [15,16]. The sample was selected at random from the polling lists available in each town hall. Response rate reached 60 of the people contacted. Subjects were screened in a health center administered by the social security system. From the initial sample of 1183 people, 182 15.4 Corresponding author. Tel.: + 33-5-61521870; fax: + 33-5- 62264240. E-mail address : ruidavetcict.fr J.-B. Ruidavets. 0021-915000 - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 0 0 0 0 3 7 7 - 4 consecutive subjects 100 men and 82 women were included in the specific study to assess + -catechin concentration in plasma in relation to dietary patterns. Participants were volunteers. Subjects were informed of the aim of the study and a formal consent form was completed and signed by each subject. Authorization from the appropriate Ethics Committee was obtained. Subjects were contacted by letter. Those who agreed to participate were given a morning appointment and asked to fast for a period of 10 h minimum. No food and no drink except water were allowed during the fasting period. 2 . 2 . Anthropometric, clinical measurements and medical history Subjects were screened in the morning for cardiovas- cular risk factors by trained nurses. Anthropometric measurements including weight, height, waist and hip circumferences were measured with standardized proce- dures [17]. Body mass index BMI and waist-to-hip ratio WHR were calculated as follows, weight kg height m 2 and waisthip. Blood pressure was mea- sured twice after a 5-min rest at least in a sitting position using a standard mercury sphygmomanometer. The average of the two blood pressure measurements was used for the analysis. Questionnaires were administered by interviewers at the examination center. Information on demographic and socio-economic factors, educational level, occupa- tional activity, personal previous history, tobacco con- sumption, drug intake, physical activity and medical problems was collected. 2 . 3 . Dietary measurements Subjects were asked to have their evening meal be- tween 19:00 and 23:00 h. They were asked not to eat or drink except water after 23:00 h and to comply with a minimum fasting period of 10 h before the blood collection on the next morning between 08:00 and 10.30 h. Subjects were instructed not to change anything in their nutritional habits. The average fasting period was 11 h and 55 min ranging from 10 to 15 h. Food intake was assessed using a food recall method to test the relationship between dietary catechin intake and + - catechin concentration in plasma. After blood collec- tion, subjects were interviewed concerning their last evening meal by a qualified dietitian. Food and drink consumption was carefully verified by the dietitian who, in order to avoid forgotten or misreported data, sub- mitted a list of various food categories to the subjects to check the reliability of the data. Food intake was assessed using estimated weights of food, household measures and portion size. The composition of pre- pared dishes was written down and the actual quantity consumed calculated according to the number of people sharing the meal, allowing for different portion sizes. For meals, which were not taken at home, the cook was contacted, the composition of the recipes and the por- tion size were recorded. The estimates were facilitated by the use of photographs showing portion sizes and their respective weights. Among the 182 selected sub- jects, two were excluded from the statistical analysis because they had tea during their last meal the day before and the number of subjects was too small to be analyzed. Food data were translated into nutrients us- ing Renaud food composition table [18]. A catechin- food composition table was created using a compilation of the results of the measurements performed for differ- ent plant food and alcoholic beverages already pub- lished in some reports and in some food-composition tables [19 – 23]. This table Appendix A gives the cate- chin average values supplied by 100 g of food. It allowed the calculation of the subjects’ total catechin intake, taking into account the subjects’ specific food intake. To check that the dietary pattern of the evening meals was comparable with their usual nutritional habits, a nutritional study using a 3-day food record method [24] was carried out among a third of the original sample n = 56. Moreover, to check that the alcohol intake during the subjects’ last evening meal was in agreement with their usual drinking habits, a quantitative recall frequency questionnaire was filled in, specifying for each day of the week, each type of alcoholic beverage intake wine, beer, spirits, cider or aperitifs representative of the last 12-month consumption. 2 . 4 . Blood sample collection and biological analysis A blood sample was collected after a 10-h-fasting period minimum, in tubes containing Na 2 EDTA, kept at room temperature and centrifuged within 3 h. Plasma tubes were stored at − 80°C temperature. Plasma total cholesterol and triacylglycerol were mea- sured by enzymatic method. High-density cholesterol was measured in the supernatant after phospho- tungstatemagnesium chloride precipitation. + -Cate- chin was analyzed in plasma by high-performance liquid chromatography HPLC method using fluores- cence detection. Sample preparation included precipita- tion of plasma proteins using acetonitrile, followed by direct injection into the HPLC system using fluores- cence detection. A Hewlett-Packard, Model 1090 HPLC system with three low pressure pumps and diode array UV detector coupled to a Hewlett-Packard Chemistation was used for solvent delivery and detec- tion. A Shimadzu RF 530 fluorescence detector coupled with a Safram recorder was also used for detection. A Hewlett-Packard column Nucleosil 100 C 18 , 250 × 4 mm, 5 mm particle size was used as the stationary phase. The solvents used for separation were A, 50 mM dihydrogen ammonium phosphate adjusted to pH 2.6 with orthophosphoric acid; B, 20 A with 80 acetoni- trile; C, 0.2 M orthophosphoric acid adjusted with ammonia to pH 1.5. The column was thermostated at 25° and the flow-rate was 0.7 mlmin. Fluorescence detection was monitored at an excitation wavelength of 280 nm and an emission wavelength of 310 nm. A 0.5-ml aliquot of the plasma sample was placed in an ice bath. Acetonitrile 0.7 ml was added to effect protein precipitation. The mixture was then centrifuged at 2000 × g and + 4°C for 4 min. The supernatant was taken into a 1-ml syringe and filtered through a 0.5 mm pore size membrane. A 25 ml volume of filtered super- natant was directly injected into the HPLC system. A stock solution of + -catechin was prepared by dis- solving 20 mg of the compound in 10 ml of methanol. It was stored at + 4°C, and diluted with methanol before use. The stock solution was stable for at least 1 month at + 4°C. Calibration curves were constructed by linear regression of the peak-area ratio versus con- centration. Validation of accuracy and precision were satisfactory for both within and between batch assays. With fluorescence detection, coefficients of variation were less than 6.5 and mean relative errors were within 9 4.8. The average recovery was 85.3. The limit of detection and quantification was 40 ngml [25]. Identification of catechin in plasma by mass spec- trometer was realized after collection of the HPLC peak, dry evaporation and adjustment for methanol. The sample was introduced in an El-Mass Spectrometer HP5989 via a particle beam interface with an electron energy of 70 eV. Comparison was made with a + - catechin reference standard from Aldrich F.W. 290.28. The spectrum of the peak compound identified by EI-SM was similar to the spectrum of the catechin reference standard. The levels of + -catechin in plasma did not include methylated, sulfated and glu- curonide metabolites. 2 . 5 . Definition of types of diet Up to now, various works have been interested in the pharmacokinetic aspects of some polyphenols and more particularly catechin. Their aim was to study the rela- tionship between specific food-catechin intake, such as wine or tea, and its levels found in plasma or urine. Actually, under ordinary dietary conditions, catechin bioavailability and more particularly its absorption process may depend not only on the quantity ingested but also on the composition of its various nutriments constituting the alimentary bolus. In our region, dietary catechin originates mainly from fruit, vegetables and wine. Therefore, the first aim of the study was the comparison of plasmatic catechin levels in relation with four types of diets, each one based upon the intake or not of these nutriments. Despite the fact that tea is an important source of catechin, a specific group could not be created since the number of tea-drinkers was too small. The two tea-drinkers were thus excluded from the analysis. Four types of diets were determined. Diet type I, subjects who had consumed meat, eggs, fish, dairy products or cereals during their last evening meal but no fruit, no vegetable, no wine and no tea. Diet type II, nutriments that could be found in diet type I and including fruit and vegetable consumption and no wine intake. Diet type III, nutriments that could be found in diet type I and including wine intake. Diet type IV, nutriments that could be found in diet type I, and including fruit, vegetables and wine intake. This classification into four groups may appear artifi- cial and arbitrary but it does correspond to actual nutritional behaviors. Diet type IV could be classified as ‘Mediterranean diet’. Diet type II was identical but excluding wine consumption. Lastly, diet type I was rich in food from animal origin meat, fish and dairy products, thus comparable to diet type III, which besides, included wine intake. The average consumption of red wine represented 87.5 of the total wine intake, whereas white and rose´ wines represented 2.9 and 9.6 of the total wine con- sumption, respectively. 2 . 6 . Statistical methods Statistical significance of the differences between diets were tested by the x 2 -test for qualitative variables and the ANOVA test for continuous ones. An ANCOVA statistical analysis was performed to compare mean values of plasma concentration of catechin in the four types of diet after adjustment for confounding factors such as sex, age, smoking habits, BMI and WHR. A general linear model was made to test the contribution of different food intake to the level of the plasma concentration of + -catechin. For variables with a skewed distribution, analysis was performed after loga- rithmic transformation. When a distribution approxi- mating to normal was not obtained after the logarithmic transformation, a Kruskal – Wallis test was performed. Post hoc tests were done when the overall test was significant P B 0.05. Scheffe test was per- formed after ANOVA and Mann – Whitney test after Kruskal – Wallis test. Mann – Whitney significant level was adjusted for the number of pairs. Pearson correla- tion was performed to test the relationship between dietary catechin and plasma concentration of + -cate- chin. Statistical analysis was conducted using the SAS statistical system [26].

3. Results