Estimated dietary polyphenol intake and major food sources of the

Estimated dietary polyphenol intake and major food sources of

the Brazilian population

Renata A. Carnauba

*, Neuza M. A. Hassimotto

and Franco M. Lajolo

Department of Food Science and Experimental Nutrition, School of Pharmaceutical Sciences, University of São Paulo, São

Paulo, Brazil

Food Research Center, CEPID-FAPESP (Research Innovation and Dissemination Centers, São Paulo Research Founda tion),

São Paulo, Brazil

(Submitted 30 August 2020 – Final revision received 8 October 2020 – Accepted 20 October 2020 – First published online 27 October 2020)

Abstract

Epidemiological studies have suggested an inverse association between polyphenol intake and the risk of chronic diseases. However, the lack of

comprehensive data on population-level intakes, especially in Latin American countries, has limited research on this topic. We aimed to estimate

total and individual polyphenol intakes and determine the major dietary contributors in a representative sample of the Brazilian population.

Data were obtained from the Brazilian Household Budget Survey, a cross-sectional survey which included data on individual food intake of

34 003 subjects aged 10 years and over collected using two 24-h dietary records. Polyphenol content of foods was identified using the Phenol-

Explorer database and Brazilian Food Composition Database. Total and individual polyphenol intake was calculated, as well as the intake

distribution by socio-demographic factors. The median and 25– 75th percentiles of polyphenol intake were 364·3 and 200·9–1008 mg/d, respec-

tively. After energy adjustment, the median and 25–75th percentiles of polyphenol intake were 204 and 111·4–542·1 mg/1000 kcal/d (4184 kJ/d),

respectively. Non-alcoholic beverages and fruits were the major polyphenol suppliers, and coffee and orange juice the main individual food

contributors to polyphenol intake. The individual compounds most consumed were isomers of chlorogenic acid (5-caffeoylquinic acid,

4-caffeoylquinic acid, 3-caffeoylquinic acid), naringenin and hesperetin. The present study provides, for the first time, data on dietary intake

of total and individual polyphenols by the Brazilian population and illustrates the low quality of their diet. These results will facilitate the study

of associations between polyphenol class intake and health outcomes, and will also be useful for future dietary intake recommendations.

Key words: Polyphenols: Phenolic acids: Flavonoids: Stilbenes: Lignans: Brazilian population

Polyphenols are a large and complex group of plant secondary

metabolites present in plant-based foods and beverages, such

as fruits, vegetables, cereals, legumes, nuts, cocoa, wine, coffee

and tea. Over 500 different molecules are known in foods, which

can be divided into classes and subclasses according to their

chemical structure, ranging from low molecular weight to highly

polymerised compounds. The four main polyphenol classes are

phenolic acids, flavonoids, stilbenes and lignans and occur either

as glycoside, ester or polymer, or in the free form (aglycone)

(1)

Cohort studies have suggested an inverse association

between polyphenol intake and the risk of chronic diseases,

such as CVD

(2,3)

, type 2 diabetes

(4,5)

, some types of cancer

(6–8)

and neurodegenerative disease

(9)

, and overall mortality

(10)

However, the glycosylation pattern, hydroxylation, acylation,

esterification and degree of polymerisation have a major influ-

ence on the bioavailability and bioactivity of dietary poly-

phenols; thus, it is important to investigate the intake of

individual polyphenols and their role in health outcomes and

disease prevention. Assessing dietary polyphenol intake and

their major food sources is the first step towards documenting

associations between polyphenols and diseases

(11)

. In this

regard, several databases on polyphenol content in foods have

been used to assess the intake in populations, and the Phenol-

Explorer database (www.phenol-explorer.eu/) is a useful tool

for estimating individual polyphenol intake as it provides data

on 501 different polyphenols in 452 foods

(12)

To date, studies assessing the total, classes and individual

polyphenol intake have been published

(13–17)

; however, most

of them are developed in European countries

(14,15,17–25)

. More

studies in different geographical areas are needed, such as

Latin America, because distinct dietary habits are observed

and the amount (and diversity) of polyphenols consumed may

vary immensely. Few data are available for Latin American coun-

tries, except for Mexico

(26)

and Argentina

(27)

. Previous studies

Abbreviations: HBS, Household Budget Survey; RF, retention factor.

* Corresponding author: Renata A. Carnauba, email [email protected]

British Journal of Nutrition (2021), 126, 441–448 doi:10.1017/S0007114520004237

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press

were conducted with the Brazilian population, but these

included small samples

(28,29)

or the estimates were based on

non-specified food consumption data

(30)

. There are no data on

the estimated dietary intake of total and individual polyphenols

in a representative sample of the Brazilian population. Thus, the

present study aimed to estimate dietary intake of total and indi-

vidual polyphenols in a representative sample of the Brazilian

general population aged 10 years or older.

Methods

Study population

The data of the present study were retrieved from the Brazilian

National Dietary Survey, conducted along with the 2008–2009

Household Budget Survey (HBS). The HBS is a survey carried

out by the Brazilian Institute of Geography and Statistics and

designed to assess Brazilian family’s consumption structures.

In summary, the 2008–2009 HBS used two-stage cluster sam-

pling involving census sectors stratified according to different

geographical areas and socio-economic classes. First, the sectors

were randomly selected from each stratum with probability pro-

portional to the number of households present and, in the sec-

ond stage, the households were selected by simple random

sampling. Of the 55 970 households sampled from the 2008–

2009 HBS, a subsample of about 25 % of households was ran-

domly selected for data collection on individual food intake

(see online Supplementary Fig. S1). Thus, the final sample

included 13 596 households, which corresponded to 34 003 sub-

jects ≥10 years old

(31)

. The research protocol was approved by

the Ethics Committee of the Institute of Social Medicine of the

State University of Rio de Janeiro (CAAE 0011.0.259.000-11).

Dietary intake

Dietary intake was evaluated during the 12 months of research to

encompass seasonal variations in food intakes. Individual food

intake data were obtained using two 24-h dietary records on

non-consecutive days. Participants were instructed to record all

food and beverages consumed, including amount (household

or volume measurements), food preparation technique (e.g. bak-

ing, broiling, frying and roasting), time and place of consumption

(inside or outside the home). An instruction manual including

photographs to facilitate the estimation of portion sizes was used.

At the end, trained professionalsreviewed the dietary records with

the participant to prevent that any food was forgotten.

Recipes were converted into ingredients to estimate the

amount of all ingredients in each mixed dish using the software

‘ILSI: Analysis of the average food consumption per capita of the

Brazilian population’

(32)

, which contains information regarding

the amounts of each food that composes the recipes mentioned

in the 2008–2009 HBS dietary records.

Correspondence between food items in dietary records

and in food composition database on polyphenols

For the present study, the dietary polyphenol intake was esti-

mated from the two dietary records. The Phenol-Explorer data-

base (www.phenol-explorer.eu/) was used to obtain data on the

polyphenol content in foods, which contain data on 501 different

polyphenols divided into eighteen polyphenol classes, such as

phenolic acids, flavonoids, stilbenes and lignans. All individual

polyphenols were determined by reverse-phase HPLC, except

for proanthocyanidins, data of which correspond to normal-

phase HPLC. For foods that contained polyphenols linked to

the food matrix and only solubilised and quantified after basic

or acid hydrolysis, content values were obtained by chromatog-

raphy after hydrolysis, such as lignans in all foods, ellagic acid in

walnuts and hydroxycinnamic acids in cereals, legumes and

olives

(12)

. For some regional foods and others highly prevalent

in the Brazilian diet (e.g. beans and orange), we used data from

foods collected and analysed in Brazil by HPLC. We also used

data from Brazilian Food Composition Database (TBCA)

(www.tbca.net.br/), a database which contains data on flavo-

noid content in foods harvested in Brazil and analysed by

HPLC

(33)

. When data were overlapped between the TBCA and

Phenol-Explorer database, TBCA data had priority.

The total polyphenol intake was estimated based on the

ingredient content in each mixed dish. The effects of food

processing on polyphenol content were estimated using reten-

tion factors (RF) included in the Phenol-Explorer database.

The RF estimates the polyphenol content in processed foods

by taking into account weight loss or water gain during

cooking

(34)

Estimation of polyphenol intake and dietary contributors

to polyphenol intake

The individual polyphenol intake from each food was calculated

by multiplying the content of each polyphenol by the daily

amount of each food consumed. Total polyphenol intake was

calculated as the sum of all individual polyphenol intakes from

all food sources reported in the dietary records. For those poly-

phenols that are present in foods as glycosides and esters forms,

the polyphenol intake was also calculated as aglycone equiva-

lents by removing, for each individual polyphenol, the contribu-

tion to molecular weight of the non-phenolic part of the

molecule.

Socio-demographic information

Socio-demographic characteristics were collected by trained

professionals in a face-to-face interview using a structured ques-

tionnaire and included the following information: sex, age, race,

Brazilian region, household area (urban or rural) and educa-

tional level. Brazilian region was divided into five categories:

North (it comprises the states of Acre, Amapá, Amazonas,

Pará, Rondônia, Roraima and Tocantins), Northeast (it comprises

the states of Alagoas, Bahia, Ceará, Maranhão, Paraíba, Piauí,

Pernambuco, Rio Grande do Norte and Sergipe), Middle-West

(it comprises the states of Goiás, Mato Grosso and Mato

Grosso do Sul), Southeast (it comprises the states of Espírito

Santo, Minas Gerais, Rio de Janeiro and São Paulo) and South

(it comprises the states of Paraná, Rio Grande do Sul and

Santa Catarina). Educational level was divided into three

categories: elementary school, secondary school and university.

Personal income was assessed considering the monthly per cap-

ita income of all monetary and non-monetary sources of income

442 R. A. Carnauba et al.

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press

and was divided into three categories: low income (first tercile),

middle income (second tercile) and high income (third tercile).

Statistical analyses

Data are presented as medians for continuous variables and

frequencies or percentages for categorical variables.

Polyphenol intake distribution was analysed using the

Kolmogorov–Smirnov test, and it did not follow a normal distri-

bution. Thus, dietary polyphenol intakes were presented as

medians and 25th and 75th percentiles, and the Kruskal–

Wallis test with Dunn post hoc (significance values adjusted by

Bonferroni correction) was used to test differences according

to socio-demographic characteristics. Because the polyphenol

intake increases with a higher food consumption, the poly-

phenol intake was also calculated in energy-adjusted terms

(mg of polyphenols per 1000 kcal/d (4184 kJ/d) of total energy

consumed). The main food contributors to the intake of

polyphenols were determined by percentage of contribution.

Estimates were performed using sample weights to allow

population representativeness, with significance level of 5 %

(P ≤ 0·05), using Stata software version 14.

Results

Total polyphenol intake

Information on 34 003 participants was available for analysis,

comprising 15 700 men and 18 303 women. A total of 1971 food

items were mentioned in the food records. After converting rec-

ipes into ingredients and excluding repeated foods, a list of 417

different food items was obtained and 229 have polyphenol con-

tent described (177 from Phenol-Explorer database and fifty-two

from TBCA and published Brazilian data). Also, 253 polyphenols

were found in these foods in aglycone, glycoside and esters

forms, which correspond to 124 different aglycone polyphenols

(see online Supplementary Table S1).

The median and 25–75th percentiles of polyphenol intake as

aglycone equivalents for the whole population were 364·3 and

200·9–1008 mg/d, respectively. After energy adjustment, the

median and 25–75th percentiles of polyphenol intake were

204 and 111·4–542·1 mg/1000 kcal/d (4184 kJ/d), respectively.

The energy-adjusted polyphenol intake as aglycone equivalents

according to socio-demographic characteristics is presented in

Table 1. The intake of total polyphenols, phenolic acids and fla-

vonoids was higher among females than in males (all P = 0·0001)

after adjusting for energy intake (see online Supplementary

Table S2). The consumption of total polyphenols, phenolic acids

and lignans increased with age, unlike flavonoids, whose highest

median intake was observed for adolescents (14–18 years). Total

polyphenols intake was significantly higher in South (218 mg/d),

compared with other regions, and lower in Midwest (171 mg/d).

Phenolic acid intake was greatest in Northeast (95 mg/d), and

flavonoid intake was highest in South (82 mg/d). Subjects living

in rural areas consumed more total polyphenols, phenolic

acids and lignans compared with those living in urban areas

(P = 0·0001). In contrast, the median intake of flavonoids and

other polyphenols was higher in the urban area compared with

the rural area (P = 0·0001).

Subjects with lower educational level (elementary school)

showed a higher phenolic acid and lignan intake than did sub-

jects with higher educational level (university) (P = 0·0001).

Additionally, subjects in lowest personal income tercile showed

higher phenolic acids intake compared with subjects with higher

personal income (P = 0·0001). The opposite was found for flavo-

noid intake, with higher median intake for subjects in the highest

educational level (compared with lowest educational level

group, P = 0·0001) and personal income category (compared

with lowest personal income category, P = 0·0001).

The polyphenol classes intake and the contribution of indi-

vidual polyphenols to the class are shown in Table 2.

Hydroxycinnamic acids were the most consumed polyphenol

class (299·9 and 152·8 mg/d as aglycone equivalents). The sec-

ond most abundant polyphenol class consumed were flava-

nones (86·5 and 58·9 mg/d as aglycone equivalents). Other

classes contributed to a lesser extent to the polyphenol intake.

Flavones and flavan3ols were the two next classes with higher

intakes, and it was estimated in 12·8 and 7·4 mg/d (6 and

6·3 mg/d as aglycone equivalents, respectively). Other poly-

phenols included alkylphenols (1·1 mg/d), lignans (1 mg/d),

tyrosols (0·8 mg/d), alkylmethoxyphenols (0·2 mg/d), stilbenes

(0·003 mg/d), furanocoumarins, hydroxybenzaldehydes,

hydroxycoumarins, phenolic terpenes, catechols, pyrogallols

and phlorins (0·1 mg/d).

The individual polyphenols with higher contribution to total

polyphenol and hydroxycinnamic acids intake were isomers of

chlorogenic acid (5-caffeoylquinic acid, 4-caffeoylquinic acid

and 3-caffeoylquinic acid). Naringenin, hesperetin and naringe-

nin 7-O-glucoside were the main individual flavanones con-

sumed, with contribution of 42·2, 21·2 and 20·1 % to the

subclass intake, respectively.

Polyphenol intake per food groups and main food

contributors

The main food group sources as well as the three highest food

contributors to total, classes and subclasses of polyphenol intake

are shown in Table 3. Non-alcoholic beverages and fruits were

the main polyphenol suppliers (contributions of 76 and 13 %,

respectively), followed by cereals (3·5 %) and vegetables

(3·1 %). The main individual food contributors to polyphenol

intake were coffee and orange juice, with respective contribu-

tion of 59·4 and 16·3 % to total polyphenols intake. Non-

alcoholic beverages and vegetables were the main food group

for phenolic acids and hydroxycinnamic acids, with coffee,

potato and bean as the main individual food contributors to

the class and subclass intake. Non-alcoholic beverages, fruits

and cocoa products were the main flavonoid providers with

respective contribution of 58, 28·9 and 4·5 %.With respect to fla-

vonoid subclasses, non-alcoholic beverages and fruits were the

most abundant food group to flavanone intake (contribution of

76 and 24 %, respectively) and coca products and fruits were the

richest food group to flavan3ol intake (66·8 and 14·9 %, respec-

tively). Orange juice and orange were the main individual food

contributors to total flavonoid and flavanone intake.

Polyphenol intake of Brazilian population 443

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press

Table 1. Energy-adjusted total and polyphenol class intake (mg/1000 kcal/d (4184 kJ/d)) by socio-demographic characteristics*†

(Median values and 25–75th percentiles)

Total polyphenols Phenolic acids Flavonoids Lignans Stilbenes Other polyphenols‡

Median

25–75th

percentiles Median

25–75th

percentiles Median

25–75th

percentiles Median

25–75th

percentiles Median

25–75th

percentiles Median

25–75th

percentiles

Sex

Male 15 700 194 105–515 81 39–143 57 15–399 0·60·4–10·0002 0·0–0·001 1 0·6–1·4

Female 18 303 214 118–568 86 43–151 71 20–445 0·50·3–0·90·0003 0·0–0·001 1 0·7–1·5

P 0·0001 0·0001 0·0001 0·0001 0

·0001 0·0001

Race

White 13 705 205 113–566 77 38–139 74 21–459 0·60·3–10·0003 0·0–0·001 1·10·7–1·5

Others 20 298 203 110–521 89 44–157 54 15–383 0·60·4–10·0002 0·0–0·001 1 0·6–1·4

P 0·0001 0·0001 0·0001 0·0152 0·0002 0·0001

Age (years)

10–13 3084 173 96–488 56 19–115 70 23–385 0·50·3–0·80·0003 0·0–0·001 1 0

·7–1·4

14–18 3855 187 100–522 60 21–121 74 23–434 0·50·3–0·90·0003 0·0–0·001 1 0·7–1·4

19–59 22 742 205 113–550 86 45–148 63 16–428 0·60·3–10·0003 0·0–0·001 1 0·7–1·5

≥60 4322 225 129–565 108 58–181 55 16–413 0·70·4–1·10·0002 0·0–0·001 1 0·6–1·4

P 0·0001 0·

0001 0·0001 0·0001 0·1011 0·0001

Brazilian region

North 5274 202 101–539 69 36–122 76 17–441 0·40·2–0·80·0002 0·0–0·001 0·80·5–1·2

Northeast 12 615 199 106–475 95 48–172 42 14–311 0·60·3–0·90·0003 0·0–0·001 0·90·5–1·4

Midwest 4645 171 95–532 67 32–121 63 16–440 0·60·4–1·00·0002 0·0–0·001 0·90·6–1

·3

Southeast 7302 210 117–565 84 40–143 72 19–448 0·60·4–1·00·0002 0·0–0·001 1 0·7–1·4

South 4167 218 118–634 77 39–140 82 24–524 0·60·3–1·10·0003 0·0–0·001 1·20·9–1·7

P 0·0001 0·0001 0·0001 0·0001 0·0001 0·0001

Area

Urban 25 753 202 111–552 80 39–141 67 18–442 0·60·3–0·90·0003 0·0–0·001 1·

10·7–1·5

Rural 8250 215 114–510 102 53–178 46 14–341 0·70·4–1·10·0002 0·0–0·001 0·70·4–1·2

P 0·1017 0·0001 0·0001 0·0001 0·0001 0·0001

Educational level

Elementary school 3966 210 116–506 102 57–171 48 14–341 0·70·4–1·00·0002 0·0–0·001 0·60·6–1·4

Secondary school 5138 207 111–603 76 38–133 75 20–497 0·60·3–0·90·0003 0

·0–0·001 1·10·7–1·5

University 1705 215 122–650 67 36–119 107 37–556 0·50·3–0·90·0003 0·0–0·001 1·20·8–1·7

P 0·0032 0·0001 0·0001 0·0001 0·0001 0·0001

Personal income

Low 11 335 207 109–478 100 51–177 43 14–312 0·60·4–1·00·0002 0·0001–0·001 0·90·5–1·3

Middle 11 335 196 107–512 87 46–154 52 15–383 0·60·

4–1·00·0003 0·0001–0·001 1 0·6–1·4

High 11 333 208 116–607 72 33–126 87 25–521 0·50·3–0·90·0003 0·0001–0·001 1·10·8–1·6

P 0·0014 0·0001 0·0001 0·0001 0·0001 0·0001

* Estimates were performed using sample weights to allow population representativeness.

† Comparisons across categories were performed by using the Kruskal–Wallis test.

‡ Other polyphenols as the sum of alkylphenols, alkylmethoxyphenols, furanocoumarins, hydroxybenzaldehydes, hydroxycoumarins, phenolic terpenes, tyrosols, catechols, pyrogallols and phlorins.

444 R. A. Carnauba et al.

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press

In regard to other polyphenol classes, vegetables and fruits

were the main food group for lignan intake, with cabbage and

orange as the major individual food contributors to the class

intake (contributions of 35·5 and 17·4 %, respectively). Fruits

were the main source for stilbenes intake, and cereals the major

food group for other polyphenols (mainly alkylphenols) intake.

Table 2. Main polyphenol subclass intake (mg/d) and the three most consumed individual polyphenols

(Median values and 25–75th percentiles)

Polyphenol classes

Total in the subclass Total as aglycone

Top three most consumed individual

polyphenolsMedian 25–75th percentiles Median 25–75th percentiles

Hydroxybenzoic acids 0·80·5–1·10·80·5–1·1 4-Hydroxybenzoic acid (37·6 %), vanillic

acid (37·6 %), syringic acid (11·8%)

Hydroxycinnamic acids 299·9 145·7–534·5 152·875·7–269·3 5-Caffeoylquinic acid (34 %),

4-Caffeoylquinic acid (26·9 %),

3-Caffeoylquinic acid (23·2%)

Anthocyanins 0·50·02–0·70·07 0·01–0·5 Cyanidin (50 %), malvidin 3-O-glucoside

(20 %), delphinidin (10 %)

Flavan3ols* 7·42·2–29·46·31·6–28 Epicatechin (68·5 %), catechin (20 %),

gallocatechin (7·2%)

Flavanones 86·58·8–847·658·96·1–796·5 Naringenin (42·2 %), hesperetin (21·2 %),

naringenin 7-O-glucoside (20·1%)

Flavones 12·88·1–18·96 3·9–8·6 Apigenin 6,8-C-galactoside-C-arabinoside

(48·6 %), apigenin 6,8-C-arabinoside-C-

glucoside (32·4 %), luteolin (9·9%)

Flavonols 6·12·1–8·63·61·7–6·9 Quercetin (55 %), quercetin 4-O-glucoside

(29·2 %), kaempferol (7·4%)

Other polyphenols 3·42·3–5·13·22·2–4·5 5-Heneicosylresorcinol (13·8 %),

secoisolariciresinol (12·3 %), lariciresinol

(6·4%)

Total 680·1 313

·2–1093·5 364·3 200·9–1008 5-Caffeoylquinic acid (22·4 %),

4-caffeoylquinic acid (17·7 %),

3-Caffeoylquinic acid (15·3%)

* Includes proanthocyanidins.

Table 3. Contribution (%) of foo d groups to total, classes and subclasses of polyphenol intake

Non-alcoholic

beverages

Alcoholic

beverages Fruits Vegetables Cereals

Oils, seeds

and nuts

Cocoa

products

Other

foods

Main food contributors (%

contribution to polyphenol class)

Phenolic acids 91 0·80·53·92·50·50·50·3 Coffee (90·6 %), potato (2·5 %), bean

(0·7%)

Hydroxybenzoic

acids

0·0120·028 55 5 0·00·0 Rice (54·7 %), beer (8 %), wine (4 %)

Hydroxycinnamic

acids

92 0·20·53·82·20·40·50·4 Coffee (91· 8 %), potato (2·6 %), bean

(1 %)

Flavonoids 58 0·328·92·53·50·74·51·6 Orange juice (57·2 %), orange

(15·8 %), chocolate (4·7%)

Flavan3ols 4·71·514·94·

30·07·866·80·0 Chocolate (51·3 %), chocolate

powder (14 %), banana (2·3%)

Flavones 19·30·04·62·860·60·01·411·3 Bread (31·7 %), orange juice

(17·8 %), cracker (2 %)

Flavonols 8 2 8·563·56 0·00·012·0 Kale (28 %), bean (11·3 %), apple

(7·6%)

Flavanones 76 0·024 0·00·00·00·00·0 Orange juice (75 %), orange (21 %),

lemon (0·5%)

Anthocyanins 0·08·591·50·00·00·00·00·0 Açaí (71 %), grape (17 %), wine

(8·5%)

Lignans 0·00·017·569·510 3 0·00·0 Cabbage (35

·5 %), orange (17·4 %),

nuts (3 %)

Stilbenes 0·05·888·75·50·00·00·00·0 Grape (86·5 %), peanut (5·4 %), wine

(5 %)

Other

polyphenols

8·411·70·62 72·35 0·00·0 Wheat flour products (67·7 %), beer

(11·4 %), coffee (8·1%)

Total polyphenols 76 0·313 3·13·50·72·20·9 Coffee (59·4 %), orange juice

(16·3 %), orange (4·5%)

Polyphenol intake of Brazilian population 445

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press

Discussion

To our knowledge, this is the first study to describe a detailed

intake of total polyphenols and polyphenol classes in a nation-

ally representative sample of the Brazilian population. Previous

studies focused on specific polyphenol class (flavonoid) with a

non-specific food consumption data

(30)

or small samples that do

not represent the entire population

(28,29)

. In comparison with the

available data, the present estimations of total polyphenols

intake were lower than that previously reported for elderly sub-

jects of Viçosa (533 mg/d)

(29)

. The median intake of total poly-

phenols and polyphenol classes in the present study is also

lower than values reported by other studies. French adults in

the SU.VI.MAX cohort had a mean polyphenol intake of

820 mg/d

(14)

, adults in the Finnish cohort FINDIET had an intake

of 863 mg/d

(15)

and adult females in UK had a mean intake of

1080 mg/d

(18)

(all values expressed as aglycone equivalents).

In regard to the polyphenol classes, the median flavonoid intake

found in our study is also lower than reported intakes for adults

in the USA

(13)

and Korea

(16)

. The Brazilian median intake of phe-

nolic acids and stilbenes was lower than that reported for adults

from the UK

(18)

and in the European Prospective Investigation

into Cancer and Nutrition (EPIC) cohort

(20)

The differences between studies could be explained by dif-

ferent dietary preferences and habits according to different pop-

ulations, which are related to the food availability and local

cultures. Specifically, in our analysis, it is important to highlight

some considerations about the Brazilian dietary pattern, which

can justify the low polyphenol intake. During the last decades,

there was an increase in processed food intake and replacement

of traditional meals with high-energy foods, rich in fat and added

sugars. The per capita fruit and vegetable intake is very low

(69·1 g/d for men and 92·6 g/d for women), even lower than that

reported in Viçosa study (per capita fruit and vegetable intake

was 247·6 g/d). In addition, the low food variety (just 417 differ-

ent food items mentioned in dietary records) in the Brazilian diet

resulted in a reduced number of consumed polyphenols, lower

than that reported in other populations

(14,26)

. Together, these fig-

ures illustrate the low quality of the Brazilian diet and point to

important health issues, corroborating the nutritional concern

of the country.

It is also important to point that food composition data of

some local fruits and other plant-based foods mentioned in

dietary records are not available. Although they represented less

than 5 % of total fruit and vegetable intake, the systematic pub-

lication of the polyphenol food composition data, as well as

comprehensive compilation of these information, should be

encouraged to facilitate researches on this topic. Other methodo-

logical issues may explain the differences of intake between

studies, such as the use of different databases of polyphenol con-

tent in foods (which differ with respect to the polyphenol classes

analysed and to the polyphenol content in foods) and the use of

different data collection methods.

Hydroxycinnamic acids were the main contributors to total

polyphenols intake, followed by flavanones. In non-

Mediterranean

(17)

and Latin American countries, such as

Mexico

(26)

and Argentina

(27)

, although the polyphenol intake

in our study was lower in these countries, the contribution of

phenolic acids and flavonoids to the polyphenol intake was

similar. Other polyphenol classes were consumed in very low

amounts, which could be explained to their low content in foods

and to the low intake of main food sources. Although these

polyphenols classes exhibit promising biological properties,

their low daily intake by the Brazilian population and

others

(14,18,19,21–23,26)

raises issues to their importance in health

and lower disease risk.

Coffee was the main food source for total polyphenols and

hydroxycinnamic acids intake, corroborating with studies

conducted with São Paulo

(28)

, Viçosa

(29)

and other Western pop-

ulations

(18,22,23)

. Orange juice and oranges were the main

contributors to total polyphenols and flavonoid intake.

Previous studies support that orange is the main food source

for flavonoids intake by the Brazilian population

(28,30,35)

In the study conducted with Viçosa

(29)

, beans had a higher

contribution to the flavonoid intake. It is important to note that

in our study, the phenolic content in beans was extracted from a

research evaluating the phenolic profile of Brazilian bean culti-

vars after soaking and cooking

(36)

. The previous study consid-

ered data from Phenol-Explorer, and the RF information is not

available to all beans’ polyphenols. This fact may justify the

lower contribution of beans to the polyphenol intake in

our study.

In our study, total polyphenols and polyphenol classes intake

was higher in females than males after adjusting for energy

intake, suggesting that the higher intake in men is due to the

higher food intake, while the higher intake in females is due

to a higher polyphenol-rich foods intake

(24)

. Total polyphenols

and phenolic acids intake increased with age, probably due to

the greater intake of the main food contributor to polyphenol

and phenolic acids (coffee). In addition, age is inversely related

to highly processed foods intake and directly related to vegetable

consumption

(37)

Total polyphenolandpolyphenolclassesintakevariedwidely

according to country region. Total polyphenols and flavonoid

intake was higher in South (218 and 82 mg/d, respectively),

whereas Midwest had the lowest total polyphenols and phenolic

acids intake (171 and 67 mg/d, respectively). In Brazil, dietary

habits vary according to the region, and in North, Northeast

and Midwest regions, the intake of local fruits and vegetables is

higher. Tea intake is highly prevalent in South, which could

explain the higher total intake of total polyphenols and flavo-

noids

(38)

. Median intake of total polyphenols and phenolic acids

was higher in rural areas. The population of urban areas con-

sumes a higher amount of highly processed foods, which could

explain the lower medians in urban area

(36)

. However, flavonoid

intake was higher in urban areas, corroborating the previously

reported for adult females in the Mexican Teachers’ Cohort

(26)

A potential explanation could be that subjects in rural areas con-

sume more local plant-based foods, such as fruits and vegetables,

with limited polyphenol content described.

The major strengths of the present study are the large and

comprehensive sample size, added to the use of sample expan-

sion factors, which provides a reliable approximation of the sit-

uation in Brazil. Other strengths include the assessment of intake

by socio-demographic factors, which is useful for establishing

public health policies, and the inclusion of polyphenol content

446 R. A. Carnauba et al.

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press

information in Brazilian foods obtained by high-quality analyti-

cal methods, in addition to the use of the Phenol-Explorer data-

base, the most comprehensive database for polyphenols. The

separation of the ingredient content in each mixed dish and

the use of the RF is another strength, since it provides more reli-

able data. However, the present study has also certain limita-

tions. As mentioned, information on some local fruits and

vegetables is still scarce because they have not been character-

ised or only poorly characterised. Other limitation is the lack of

information about polyphenol content on cooked foods, since

RF information in the Phenol-Explorer database is not compre-

hensive to all foods, compounds and processing methods.

In conclusion, for the first time, the present study provided

data on total polyphenols and polyphenol classes intake in a

nationally representative sample of the Brazilian population.

Our results corroborate central nutritional concern in Brazil, in

which the intake of traditional meals and plant-based foods

has been decreasing in the last few decades. The results

described will facilitate the study associations between the poly-

phenol classes intake and health outcomes and will be useful to

investigate the potential role on health of certain foods con-

sumed in greater amounts by the Brazilian population, such as

coffee and orange juice.

Acknowledgements

The authors would like to thank International Life Sciences

Institute Brasil (Task Force Agrochemicals) for technical support

in the initial phase of the study. The authors thank Food Research

Center – CEPID-FAPESP (Research Innovation and

Dissemination Centers, São Paulo Research Foundation) and

the National Counsel of Technological and Scientific

Development (CNPq).

This work was not supported by any funding agency.

F. M. L. designed, planned and assisted the research, and

reviewed the manuscript; R. A. C. conducted research, per-

formed data analysis and wrote the paper; N. M. A. H. critically

reviewed the manuscript. All authors read and approved the final

version of the manuscript.

There are no conflicts of interest.

Supplementary material

For supplementary materials referred to in this article, please visit

https://doi.org/10.1017/S0007114520004237

References

1. Pérez-Jiménez J, Neveu V, Vos F, et al. (2010) Systematic analy-

sis of the content of 502 polyphenols in 452 foods and bever-

ages: an application of the Phenol-Explorer database. J Agric

Food Chem 58, 4959–4969.

2. Cassidy A, Mukamal KJ, Liu L, et al. (2013) A high anthocyanin

intake is associated with a reduced risk of myocardial infarction

in young and middle-aged women. Circulation 127, 188–196.

3. Lajous M, Rossignol E, Fagherazzi G, et al. (2016) Flavonoid

intake and incident hypertension in women. Am J Clin Nutr

103, 1091–1098.

4. Tresserra-Rimbau A, Guasch-Ferré M, Salas-Salvad´oJ,et al.

(2015) Intake of total polyphenols and some classes of poly-

phenols is inversely associated with diabetes in elderly people

at high cardiovascular disease risk. J Nutr 146, 767–777.

5. Zamora-Ros R, Forouhi NG, Sharp SJ, et al. (2013) The associ-

ation between dietary flavonoid and lignan intakes and incident

type 2 diabetes in European populations: the EPIC-InterAct

study. Diabetes Care 36, 3961–3970.

6. Zamora-Ros R, Cayssials V, JenabM, et al. (2018)Dietary intake of

total polyphenol and polyphenol classes andthe risk of colorectal

cancer in the European Prospective Investigation into Cancer and

Nutrition (EPIC) cohort. Eur J Epidemiol 33, 1063–1075.

7. Vermeulen E, Zamora-Ros R, Duell EJ, et al. (2013) Dietary fla-

vonoid intake and esophageal cancer risk in the European pro-

spective investigation into cancer and nutrition cohort. Am J

Epidemiol 178, 570–581.

8. Lin Y, Yngve A, Lagergren J, et al. (2014) A dietary pattern rich

in lignans, quercetin and resveratrol decrease the risk of oeso-

phageal cancer. Br J Nutr 112, 2002–2009.

9. Devore EE, Kang JH, Breteler MM, et al. (2012) Dietary intakes

of berries and flavonoids in relation to cognitive decline. Ann

Neurol 72, 135–43.

10. Bondonno NP, Dalgaard F, Kyro C, et al. (2019) Flavonoid

intake is associated with lower mortality in the Danish Diet

Cancer and Health Cohort. Nat Commun 10, 188–

196.

11. Wallace TC, Blumberg JB, Johnson EJ, et al. (2015) Dietary bio-

actives: establishing a scientific framework for recommended

intakes. Adv Nutr 6,1–4.

12. Neveu V, Perez-Jimenez J, Vos F, et al. (2010) Phenol-Explorer:

an online comprehensive database on polyphenol contents in

foods. Database 2010, bap024.

13. Chun OK, Chung SJ & Song WO (2007) Estimated dietary flavo-

noid intake and major food sources of U.S. adults. J Nutr 137,

1244–1252.

14. Pérez-Jiménez J, Fezeu L, Touvier M, et al. (2011) Dietary intake of

337 polyphenols in French adults. Am J Clin Nutr 93, 1220–1228.

15. Ovaskainen ML, Torronen R, Koponen JM, et al. (2008) Dietary

intake and major food sources of polyphenols in Finnish adults.

J Nutr 138, 562– 566.

16. Jun S, Shin S & Joung H (2016) Estimation of dietary flavonoid

intake and major food sources of Korean adults. Br J Nutr 115,

480–489.

17. Zamora-Ros R, Knaze V, Rothwell JA, et al. (2016) Dietary poly-

phenol intake in Europe: the European Prospective

Investigation into Cancer and Nutrition (EPIC) study. Eur J

Nutr 55, 1359–1375.

18. Yahya HM, Day A, Lawton C, et al. (2016) Dietary intake of 20

polyphenol subclasses in a cohort of UK women. Eur J Nutr 55,

1839–1847.

19. Grosso G, Stepaniak U, Topor-Madry R, et al. (2014) Estimated

dietary intake and major food sources of polyphenols in the

Polish arm of the HAPIEE study. Nutrition 30, 1398–1403.

20. Zamora-Ros R, Rothwell JA, Scalbert A, et al. (2013) Dietary

intakes and food sources of phenolic acids in the European

Prospective Investigation into Cancer and Nutrition (EPIC)

study. Br J Nutr 110, 1500–1511.

21. Tetens I, Turrini A, Tapanainen H, et al. (2013) Dietary intake

and main sources of plant lignans in five European countries.

Food Nutr Res 57.

22. Tresserra-Rimbau A, Medina-Rem´

on A, Pérez-Jiménez J, et al.

(2013) Dietary intake and major food sources of polyphenols

in a Spanish population at high cardiovascular risk: the

PREDIMED study. Nutr Metab Cardiovasc Dis 23, 953–959.

23. Wisnuwardani RW, De Henauw S, Androutsos O, et al. (2016)

Estimated dietary intake of polyphenols in European adoles-

cents: the HELENA study. Eur J Nutr 58, 2345–2363.

Polyphenol intake of Brazilian population 447

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press

24. Ziauddeen N, Rosi A, Del Rio D et al. (2018) Dietary intake of

(poly)phenols in children and adults: cross-sectional analysis of

UK National Diet and Nutrition Survey Rolling Programme

(2008–2014). Eur J Nutr 58, 3183–3198.

25. Vogiatzoglou A, Mulligan AA, Lentjes MA, et al. (2015)

Flavonoid intake in European adults (18 to 64 years). PLOS

ONE 10, e0128132.

26. Zamora-Ros R, Biessy C, Rothwell JA, et al. (2018) Dietary poly-

phenol intake and their major food sources in the Mexican

Teachers’ Cohort. Br J Nutr 120, 353–360.

27. Rossi MC, Bassett MN & Sammán NC (2018) Dietary nutritional

profile and phenolic compounds consumption in school chil-

dren of highlands of Argentine Northwest. Food Chem 238,

111–116.

28. Miranda AM, Steluti J, Fisberg RM, et al. (2016) Dietary intake

and food contributors of polyphenols in adults and elderly

adults of Sao Paulo: a population-based study. Br J Nutr 115,

1061–1070.

29. Nascimento-Souza MA, de Paiva PG, Pérez-Jiménez J, et al.

(2018) Estimated dietary intake and major food sources of poly-

phenols in elderly of Viçosa, Brazil: a population-based study.

Eur J Nutr 57, 617–627.

30. Anacleto SL, Lajolo FM & Hassimotto NMA (2019) Estimation of

dietary flavonoid intake of the Brazilian population: a compari-

son between the USDA and Phenol-Explorer databases. J Food

Compos Anal 78,1–8.

31. Instituto Brasileiro de Geografia e Estatística (IBGE) (2010)

Pesquisas de Orçamentos Familiares 2008–2009: Análise do

consumo alimentar pessoal no Brasil (2008–2009 Family

Budget Surveys: Analysis of Personal Food Consumption

in Brazil). Rio de Janeiro: Ministério do Planejamento,

Orçamento e Gestão, IBGE.

32. Pais MCN & ValerioLB(2015) Avaliação do consumo de alimen-

tos in natura da população brasileira por meio da pesquisa de

orçamentos familiares – bloco de consumo alimentar pessoal

(Evaluation of Fresh Food Consumption by the Brazilian

Population Through the Survey of Family Budgets - Personal

Food Consumption Block). São Paulo: International Life

Sciences Institute.

33. Coelho KS, Bistriche EG, Grande F, et al. (2019) 12th IFDC 2017

Special Issue – Brazilian Food Composition Table (TBCA):

development and functionalities of the online version. J Food

Compos Anal 84, 103287.

34. Rothwell JA, Perez-Jimenez J, Neveu V,

et al. (2013) Phenol-

Explorer 3.0: a major update of the Phenol-Explorer database

to incorporate data on the effects of food processing on poly-

phenol content. Database 2013, bat070.

35. Arabbi PR, Genovese MI & Lajolo FM (2004) Flavonoids in vegeta-

ble foods commonly consumed in Brazil and estimated ingestion

by the Brazilian population. J Agric Food Chem 52, 1124–1131.

36. Ranilla LG, Genovese MI & Lajolo FM (2009) Effect of different

cooking conditions on phenolic compounds and antioxidant

capacity of some selected Brazilian bean (Phaseolus vulgaris L.)

cultivars. J Agric Food Chem 57, 5734–5742.

37. Canella DA, Louzada MLC, Claro RMC, et al. (2018)

Consumption of vegetables and their relation with ultra-

processed foods in Brazil. Rev Saude Publica 52, 50.

38. Souza AM, Pereira RA, Yokoo EM, et al. (2013) Most consumed

foods in Brazil: National Dietary Survey 2008–2009. Rev Saude

Publica 47 , 190s–199s.

448 R. A. Carnauba et al.

https://doi.org/10.1017/S0007114520004237 Published online by Cambridge University Press