The diversity of the chemical structures of dietary polyphenols makes it difficult to estimate their total content in foods, and also to understand the role of polyphenols in health and the prevention of diseases. Global redox colorimetric assays have commonly been used to estimate the total polyphenol content in foods. However, these assays lack specificity. Contents of individual polyphenols have been determined by chromatography. These data, scattered in several hundred publications, have been compiled in the Phenol-Explorer database. The aim of this paper is to identify the 100 richest dietary sources of polyphenols using this database.
Advanced queries in the Phenol-Explorer database (www.phenol-explorer.eu) allowed retrieval of information on the content of 502 polyphenol glycosides, esters and aglycones in 452 foods. Total polyphenol content was calculated as the sum of the contents of all individual polyphenols. These content values were compared with the content of antioxidants estimated using the Folin assay method in the same foods. These values were also extracted from the same database. Amounts per serving were calculated using common serving sizes.
A list of the 100 richest dietary sources of polyphenols was produced, with contents varying from 15 000 mg per 100 g in cloves to 10 mg per 100 ml in rosé wine. The richest sources were various spices and dried herbs, cocoa products, some darkly coloured berries, some seeds (flaxseed) and nuts (chestnut, hazelnut) and some vegetables, including olive and globe artichoke heads. A list of the 89 foods and beverages providing more than 1 mg of total polyphenols per serving was established. A comparison of total polyphenol contents with antioxidant contents, as determined by the Folin assay, also showed that Folin values systematically exceed the total polyphenol content values.
The comprehensive Phenol-Explorer data were used for the first time to identify the richest dietary sources of polyphenols and the foods contributing most significantly to polyphenol intake as inferred from their content per serving.
Polyphenols are common antioxidants present in a large number of foods and beverages of plant origin. A mean intake of about 1 g per day has been reported (Kühnau, 1976; Ovaskainen et al., 2008). All polyphenols are characterised by the presence in their structure of one or several phenolic groups, capable of reducing reactive oxygen species and various organic substrates and minerals. These redox properties explain the considerable interest in their role in the prevention of several major chronic diseases associated with oxidative stress, such as cardiovascular diseases, cancers, type II diabetes, neurodegenerative diseases or osteoporosis (Scalbert et al., 2005).
Polyphenols show highly diverse structures and over 500 different molecules are known in foods (Neveu et al., 2010; Pérez-Jiménez et al., 2010). This diversity must be taken into account when considering bioavailability, biological properties and health effects, as the latter largely depend on their specific chemical structures (Manach et al., 2005; Loke et al., 2008). The diversity of polyphenols also makes it difficult to estimate the total polyphenol content of foods, an important fact for food and nutrition researchers and the food industry.
Ranks of foods and beverages that are richest in polyphenols have been published. For these estimations, relatively easily performed colorimetric redox assays, based on the generic reducing property of phenolic groups, have been used (Vinson et al., 1998, 1999, 2001, 2003). One of the most commonly used tests is the Folin assay (Singleton and Rossi, 1965; Scalbert, 1992). However, such global assays are prone to interference from non-phenolic food constituents. This interference may be limited by a preliminary solid phase extraction of polyphenols, but this approach is yet to be applied to a large variety of foods (Brat et al., 2006).
The United States Department of Agriculture (USDA) database, one of the most complete sources of information on the content of individual polyphenols in foods, provides composition data for 38 flavonoid aglycones in over 300 foods (USDA, 2004, 2007, 2008). This database has been used to build food composition tables for flavonoids and to estimate their intake in several cohorts (Mink et al., 2007; Ovaskainen et al., 2008). More recently, we have developed the more comprehensive Phenol-Explorer database, which includes mean content values for 502 polyphenols in 452 foods (Neveu et al., 2010; Pérez-Jiménez et al., 2010). This database not only contains data on flavonoids but also composition data for phenolic acids, lignans and stilbenes. The aim of the present work was to exploit the Phenol-Explorer data to establish a list of the 100 richest dietary sources of polyphenols and to determine the amount of polyphenols present in a food serving using common serving sizes. Polyphenol contents and amounts per serving are also compared with the amount of antioxidants in the same foods, as determined by the Folin colorimetric assay.
Materials and methods
Data on the polyphenol content in foods were extracted from the Phenol-Explorer database (www.phenol-explorer.eu). The database has been fully described elsewhere (Neveu et al., 2010; Pérez-Jiménez et al., 2010). Advanced queries in the Phenol-Explorer database allowed retrieval of information on polyphenol content, as determined by different types of analytical methods: chromatography (most often reverse-phase high-performance liquid chromatography (HPLC) and gas chromatography), normal-phase HPLC for proanthocyanidins and the Folin assay for total antioxidants. The total polyphenol content was calculated as the sum of the contents of all individual polyphenols as determined by chromatography. This includes the content of proanthocyanidin dimers and trimers. When content values for proanthocyanidins, as determined by normal-phase chromatography (dimers to polymers), were available, they were added, instead of the values obtained by reverse-phase HPLC for dimers and trimers. The main polyphenol standards used for determining Folin values in Phenol-Explorer included gallic acid, catechin and caffeic acid (59, 14 and 11%, respectively, of the values collated in the database). These three standards show similar response factors when calculated on a weight basis (Scalbert, 1992), and data are therefore roughly comparable, even if different standards were used. However, Folin values should be considered as a rough estimate of the antioxidant content because of possible differences in the response of the various antioxidants present in a food extract.
For foods containing significant amounts of polyphenols that are attached to the food matrix and can only be solubilised and quantified after basic or acid hydrolysis (phenolic acids in cereals and lignans in flaxseed) (Pérez-Jiménez et al., 2010), content values obtained by chromatography after hydrolysis were used instead of those for free polyphenols, as directly determined by chromatography. Total content of polyphenol aglycones was calculated by converting the content of each individual polyphenol in foods to that of the corresponding aglycone, taking into account the molecular weights of the glycosides or esters and aglycones. Amounts of polyphenols contained in a food serving were calculated on the basis of serving sizes published by the Food Standards Agency, UK (Food Standards Agency, 2002), the USDA (Wu et al., 2004) and by Cassidy et al. (2000) for some soy products. Some extrapolations were made for missing data; for example, in the case of a non-documented fruit juice, the serving size of a juice prepared from a related fruit was used. A retention factor was applied for cooked foods when polyphenol contents were only available for raw foods (for example, beans or pasta), to take into account the change in weight during cooking (Bognar, 2002).
Foods richest in polyphenols
The content values of all known individual polyphenols collected in the Phenol-Explorer database from the scientific literature were used to calculate total polyphenol contents of 452 foods. Data for the 100 foods with the highest polyphenol contents are shown in Table 1.
The food group with the highest number of items in this list is the seasoning group (22 items), followed by fruits (20 items), seeds (16 items), vegetables (16 items), non-alcoholic beverages (11 items), cereals (10 items), cocoa products (4 items), alcoholic beverages (3 items) and oils (2 items). Total polyphenol contents ranged from more than 15 000 mg per 100 g in cloves to 7.8 mg per 100 ml in rosé wine.
Foods richest in antioxidants
The Folin assay provides a crude estimate of the content of total antioxidants. This test is also commonly used to estimate the total polyphenol content in foods, as polyphenols with different structures have a similar response factor (Singleton and Rossi, 1965; Scalbert, 1992). However, this test is not specific for polyphenols, as the Folin reagent also reacts with other food constituents such as ascorbic acid. This assay should therefore be regarded as a global antioxidant assay, comparable to other redox assays such as the Oxygen Radical Absorbance Capacity assay. The antioxidant contents determined by the Folin assay in the same 100 foods rich in polyphenols are given in Table 1. These values are compared with the total contents of polyphenols expressed as aglycone equivalents because it is the aglycone with its phenolic group(s) that reacts with the Folin oxidant. Antioxidant contents were systematically higher than the total polyphenol contents (Figure 1).
Foods richest in polyphenols per serving
The amounts of polyphenols contained in a food serving were calculated for all foods documented in the Phenol-Explorer database. Foods containing more than 1 mg per serving are presented in Table 2. This list includes 89 foods and beverages: 23 fruits, 23 vegetables, 16 seeds, 10 non-alcoholic beverages, 6 cereals, 5 alcoholic beverages, 4 cocoa products and 2 oils. Three berries contain more than 1000 mg of polyphenols per serving. A total of 21 foods or beverages provide 100–1000 mg polyphenols per serving, 41 foods contain 10–100 mg and 24 foods 1–10 mg. Some foods rich in polyphenols, such as herbs and spices (Table 1), are absent from Table 2, as no data on serving sizes were found.
When results were expressed as aglycone equivalents (Table 2), the amount of polyphenols contained in a serving was higher than 1000 mg for 2 foods (black elderberry and black chokeberry), between 100 and 1000 mg for 22 foods, between 10 and 100 mg for 40 foods and lower than 1 mg for 25 foods.
The amount of antioxidants as determined by the Folin assay in a serving of the same foods is also shown in Table 2. Five foods contain more than 1000 mg antioxidants per serving (black elderberry, black chokeberry, globe artichoke heads, black bean and blackcurrant), 30 foods between 100 and 1000 mg and 27 foods between 10 and 100 mg.
A clear identification of the richest dietary sources of polyphenols is necessary to better understand and exploit the health-promoting effects of polyphenols, a matter of interest for the food industry, food scientists and nutritionists. However, the wide variety of chemical structures among polyphenols makes it difficult to estimate their content in foods, as different methods of analysis must be used to cover the whole range of molecules known to exist in foods. For this reason, global colorimetric assays such as the Folin assay have been used to determine the total polyphenol content in foods (Vinson et al., 1998; Vinson et al., 2001). Folin values should be considered as an estimate of antioxidant content rather than the total polyphenol content, unless applied to a purified polyphenol fraction (Brat et al., 2006). Folin values are generally well correlated to the content of antioxidants as measured by other redox assays, such as the Oxygen Radical Absorbance Capacity assay (Velioglu et al., 1998). As seen in Figure 1, the Folin values are most often higher than the total content of polyphenols, because of the presence and influence of non-phenolic antioxidants on the assay.
The most reliable method to estimate total polyphenol content is the analysis of all individual phenolic compounds. However, because of the diversity of compounds and methods needed for their analysis, the comparison of total polyphenol content in a variety of foods can only be based on the compilation of literature data. The most complete source of food composition data for polyphenols has been the USDA database for 38 flavonoid aglycones (USDA, 2004, 2007, 2008). These data have been used and subsequently augmented with other literature sources for phenolic acids to determine the total polyphenol contents in foods and to calculate polyphenol intake in cohorts (Mink et al., 2007; Ovaskainen et al., 2008). The recently developed Phenol-Explorer database provides composition data for 502 polyphenols (glycosides, esters and aglycones of flavonoids, phenolic acids, lignans, stilbenes and other polyphenols) in 452 foods. It is today the most complete database on the content of polyphenols in foods (Neveu et al., 2010). These data are used here for the first time to calculate the total polyphenol contents in these various foods and to identify the 100 richest sources of dietary polyphenols. All details on the composition of each food considered here can be obtained on the Phenol-Explorer website (www.phenol-explorer.eu; Neveu et al., 2010).
The richest sources of polyphenols are spices and herbs (Table 1). Spices such as cloves or star anise contain high amounts of phenolic flavours such as eugenol (cloves) or anethole (star anise). Dried herbs contain high amounts of flavanones such as eriocitrin (peppermint) or pinocembrin (Mexican oregano) and/or high amounts of hydroxycinnamic acids such as rosmarinic acid in herbs from the Lamiaceae family (peppermint, sage, rosemary, spearmint, thyme).
Cocoa powder and chocolate are also rich in polyphenols, mainly catechins and proanthocyanidins. A number of berries also appear at the top of the list with high contents of anthocyanins. As expected, the berries richest in polyphenols are all darkly coloured: black chokeberry, black elderberry, blueberry and blackcurrant (Table 1). Three other intensely coloured fruits follow down the list: plum, cherry and blackberry. Less-coloured fruits are found further down: strawberry, raspberry and grape, followed by pome and drupe fruits (apple, peach, apricot, nectarine, pear).
Several seeds and nuts are also found among the richest sources of polyphenols: flaxseed rich in lignan secoisolariciresinol, chestnut and walnut rich in ellagitannins, hazelnut, pecan nut and almond rich in proanthocyanidins, and soy flour and roasted soybean rich in isoflavones.
The two richest vegetable sources are black and green olives, which are particularly rich in tyrosols. They are followed by globe artichoke heads and red and green chicory (all rich in chlorogenic acid), onion (red or yellow), shallot and spinach (rich in flavonols). Other vegetables are found further down the list: broccoli, asparagus, potato, endive, lettuce, endive (escarole) and carrot.
The first beverage in the list is coffee (rich in chlorogenic acids), followed by black and green tea (rich in catechins, theaflavins and proanthocyanidins) and red wine (containing catechins, proanthocyanidins, anthocyanins and hydroxycinnamic acids). Several fruit juices are also found further down the list, containing lower amounts of polyphenols. The list also includes several cereals such as whole grain flours from wheat or rye, rich in ferulic acid and alkylresorcinols, and two plant oils (extravirgin olive oil rich in tyrosols and rapeseed oil containing 4-vinylsyringol).
The total polyphenol contents determined here using the Phenol-Explorer data can be compared with those published by Ovaskainen et al. (2008), who developed one of the most comprehensive polyphenol composition tables for 99 foods commonly consumed in Finland. A total of 33 of these foods were found among the 100 richest foods listed in this work. Polyphenol concentrations differed significantly between the two studies. In the study presented here, 11 of these 33 foods were found to have polyphenol concentrations that were more than twice as high as the Finnish values, whereas for five, the concentrations were half of those found in the Finnish list. It is not easy to explain these differences, as no details are given on the contents of individual phenolic compounds in the Finnish foods. However, some Finnish values appear surprisingly low: 1.1 mg per 100 g for onion or 2.2 mg per 100 g for brown beans, known for their high content of flavonols and flavanols, respectively (catechins and proanthocyanidins). Concentrations for other foods in the Phenol-Explorer might also be revised in the future as more data are published. For example, content values for coffee are based on no more than five original publications and could be revised as more data are published (Pérez-Jiménez et al., 2010). Furthermore, data for proanthocyanidins are still lacking for many foods, in particular fruits. This may result in a revision of the total polyphenol content of these foods as more data are published.
Some foods, such as spices and herbs, can be rich in polyphenols, but the amount consumed is low. Therefore, it is also important to estimate the amount of polyphenols in a serving (Table 2). Certain foods and beverages, such as several fruit juices and beer, do not appear in the list of foods richest in polyphenols (Table 1), but because the typical serving size is greater than 100 g, they appear in Table 2. The top 13 foods, when ranked by polyphenol concentration in a serving, are mainly fruits (seven berries and three other fruits), along with one vegetable (globe artichoke heads) and coffee. Black tea, green tea and red wine are also now higher up in the list (16th, 17th and 22nd positions). In contrast, some foods consumed in smaller amounts now appear further down the list. Dark chocolate and cocoa powder now appear at the 14th and 24th positions, instead of the 3rd and 8th positions, respectively. The same should be true for spices and herbs, although no ranking could be given because of the lack of data available on their serving size.
In summary, use of the Phenol-Explorer database on the polyphenol content of foods has established for the first time a ranking of the richest dietary sources of polyphenols based on a considerable amount of composition data published in the scientific literature. Such a ranking may be used as a reference to prioritise which foods should be studied in the future to evaluate the role of polyphenols in health and the prevention of diseases. However, limitations of the present data should not be overlooked. The total polyphenols concentration of a food is unlikely to fully explain any associated health effects because of the large differences in the bioavailability and biological activity of individual polyphenols (Manach et al., 2005; Williamson and Manach, 2005). These differences emphasise the value of using detailed content values for the 500 dietary polyphenols, as can be found in the Phenol-Explorer database.
This work was supported by Unilever, Danone and Nestlé.
About this article
Polyphenol-rich curry made with mixed spices and vegetables increases postprandial plasma GLP-1 concentration in a dose-dependent manner
European Journal of Clinical Nutrition (2018)