T Nation

Blueberries and Protein


May want to take them separately from now on...

Antioxidant activity of blueberry fruit is impaired by association with milk.
Serafini M, Testa MF, Villaño D, Pecorari M, van Wieren K, Azzini E, Brambilla A, Maiani G.

Antioxidant Research Laboratory, Unit of Human Nutrition, Istituto Nazionale di Ricerca per gli Alimenti e la Nutrizione, 00178 Rome, Italy. serafini_mauro@yahoo.it

The antioxidant properties of dietary phenolics are believed to be reduced in vivo because of their affinity for proteins. In this study we assessed the bioavailability of phenolics and the in vivo plasma antioxidant capacity after the consumption of blueberries (Vaccinium corymbosum L.) with and without milk. In a crossover design, 11 healthy human volunteers consumed either (a) 200 g of blueberries plus 200 ml of water or (b) 200 g of blueberries plus 200 ml of whole milk. Venous samples were collected at baseline and at 1, 2, and 5 h postconsumption. Ingestion of blueberries increased plasma levels of reducing and chain-breaking potential (+6.1%, p<0.001; +11.1%, p<0.05) and enhanced plasma concentrations of caffeic and ferulic acid. When blueberries and milk were ingested there was no increase in plasma antioxidant capacity. There was a reduction in the peak plasma concentrations of caffeic and ferulic acid (-49.7%, p<0.001, and -19.8%, p<0.05, respectively) as well as the overall absorption (AUC) of caffeic acid (p<0.001). The ingestion of blueberries in association with milk, thus, impairs the in vivo antioxidant properties of blueberries and reduces the absorption of caffeic acid.




LOL, I don't know if you're joking here or not. This news is actually pretty big for me, as I put 1 cup of blueberries in every single protein shake I drink.

I found an interview with a neurologist/"blueberry expert" who advises eating them a minimum of 1 hour prior to protein consumption, or 2 hours post-consumption


If 'only daily macro totals matter', I've wondered if consuming proteins, fats, and carbs separately would be beneficial, e.g. daily fats for breakfast, daily carbs pre-workout, daily protein post-workout, or something like that.
Anyone heard/read of someone trying this?


That study used milk, not just protein. Do you eat your blueberries with milk?
I'm sure there is plenty of reason to believe that protein reduces antioxidant properties of blueberries, but since the study used milk and not some sort of protein concentrate or isolate, it isn't really definitive.


I remember reading that protein also affects the benefits of green tea, although I don't recall the exact details.


Um. If that's true then it is really lame, considering cottage cheese and blueberries are a staple in my diet.


Probably the calcium in the milk. Or lactose. Or vit D etc...

Not necessarily protein or 'all' protein. Thats quite an outlandish assumption based on that study.


Wow, that sucks.....no really, i try to eat blueberries everything.
And who doesn't have a gallon of milk in the fridge?




Can anyone post access full-text article? (university students, etc). They want me to pay for it.

In the abstract, the author mentions "protein" rather than milk protein, and a review I saw of this article said the take-home message was to avoid protein (in general) within a certain time window of eating blueberries, but I won't know what they're basing this on until I see the full article


Thank you for that.


I have no idea how to post a link to it, but anyone who wants it, bust me a PM with your email address and I'll ping it through.




Terrible news.
I have blueberries with oats in the morning and with cottage cheese at night. Sigh.

How about having blueberries 10-15 minutes before protein?



I love mixing blueberries with natural yoghurt! :frowning:


If its not too much trouble, could you just copy the "methods", "results" and "discussion" parts and paste them here?


I'll grab it after. SHIT this sucks. I wonder if using soy milk would be better? I can't believe I just typed that


Edit: Christ that came out fucked up. I'll try again later.

The human diet provides a wide array of plant-derived phenolics
with antioxidant activity that helps the body to cope with oxidative
stress [1]. The biological response to antioxidant-rich foods is critically
determined by the bioavailability of active molecules. The most
abundant phenolics in the diet are not necessarily those able to reach
the highest levels in human circulation, owing to the considerable
differences in bioavailability. Human intervention studies have shown
that absorption of phenolics in the gastrointestinal tract is, typically,
between 1 and 5% of the ingested dose [2]. Polyphenols can be
absorbed in the stomach and at the small intestine level by passive
diffusion or active transport [3,4]. Once absorbed, the metabolism of
polyphenols in humans implies a profound biotransformation through
enzymatic conjugation with sulfate, methyl, or glucuronide groups in
both the small intestine epithelial cells and the liver [4]. These
metabolites can reach a maximum concentration in plasma (Cmax)1?
2 h after ingestion of phenolic-rich foods [5?7] and appear principally
in urine after 4?8h [8,9].

Variable amounts of flavonoids, not absorbed in the upper
gastrointestinal tract, reach the colon, where they are subjected to
the action of the colonic microflora, resulting in cleavage of glycosidic
linkages and the breakdown of the flavonoid heterocycle into phenolic
acids and aldehydes [10?14]. Thesemicrobial catabolites are absorbed
into the circulatory systemfromthe large intestine, reaching a Cmax ca.
5 h postconsumption and, consequently, are able to exert biological
actions [2].

Flavonoids are generally consumed in foods along with other
macronutrients such as proteins, lipids, glucose, or ethanol and these
constituents may have an impact on flavonoid bioavailability. It has
been proposed that the linking of phenolics with proteins during food
ingestion can reduce their absorption [15]. A large number of in vitro
studies have suggested that these interactions involve both the
formation of hydrogen bonds and hydrophobic interactions between
hydroxyl groups of the phenolic compounds and the carbonyl groups
of the proteins [16]. Recently, it has been shown with human
volunteers that an improvement in flow-mediated dilation, a marker
of vascular function, brought about by the ingestion of 500 ml of black
tea did not occur when the tea was consumed with milk [17].
Intervention studies with humans showed reduced bioavailability of
phenolics and in vivo antioxidant activity after the ingestion of tea and
chocolate with milk. However, other studies have failed to detect an
impact of milk and as a consequence the topic is somewhat
controversial [15,18?21] and one that requires further investigation.
In the current study, blueberries (Vacciniumcorymbosum L.), which
increase plasma antioxidant defenses and delay chemically induced
LDL and liposome oxidation ex vivo,were chosen as a dietary source of
phenolic antioxidants [22?27]. The aim of the investigation was to
Free Radical Biology & Medicine 46 (2009) 769?774
Abbreviations: AUC, area under the curve; FRAP, ferric reducing antioxidant
potential; TAC, total antioxidant capacity; TRAP, total radical-trapping antioxidant
⁎ Corresponding author. Fax: +390651494550.
E-mail address: serafini_mauro@yahoo.it (M. Serafini).
0891-5849/$ ? see front matter © 2008 Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
Free Radical Biology & Medicine
journal homepage: www.elsevier.com/locate/freeradbiomedassess whether the consumption of blueberries in association with
milk, compared with the ingestion of blueberries alone, affects in vivo
antioxidant levels and bioavailability of phenolic acids. In vitro
experiments were also carried out to investigate the effects of various
types of milk on the antioxidant capacity of a blueberry extract.
Materials and methods
Chemicals and reagents
Hydrochloric acid, methyl alcohol, and acetonitrile solvents were
purchased from Carlo Erba (Milan, Italy). Glacial acetic acid, 85% v/v
orthophosphoric acid, metaphosphoric acid, ferric chloride hexahy-
drate (FeCl3?6H2O), sodium acetate trihydrate (C2H3NaO2?3H2O), and
sodium phosphate monobase (NaH2PO4) were obtained from BDH
(Poole, England). 2,2′-Azobis (2-amidinopropane) dihydrochloride
(ABAP) was purchased from Wako Chemical (Germany), 2,4,6-tri(2-
pyridyl)-s-triazine (TPTZ) was from Fluka (Switzerland), and ferrous
sulfate hexahydrate (FeSO4?6H2O) and formic acid were from Merck
(Darmstadt, Germany). R-phycoerythrin (R-PE), phosphate-buffered
saline (PBS), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid
(Trolox), ascorbic acid, and β-glucuronidase type HP2 from Helix
pomatia were provided by Sigma?Aldrich Srl (Milan, Italy). Commer-
cial standards of gallic acid, chlorogenic acid, cyanidin 3-glucoside,
caffeic acid, ferulic acid, quercetin 3-galactoside, quercetin 3-gluco-
side, and o-coumaric acid were also from Sigma?Aldrich Srl. Double-
distilledwater (Millipore,Milan, Italy)was used throughout the study.
Food matrices
Amixture (1/1) of two varieties of blueberry (V. corymbosumL.), cv.
Berkeley and Coville, provided by the Istituto per la Valorizzazione
Tecnologica dei Prodotti Agricoli from Milan (Italy) was used. Whole
milk, semiskimmed milk, and skimmed milk were supplied by
Centrale del Latte di Roma (Rome, Italy). The nutritional composition
of the milks is displayed in Table 1.
Sample phenolic extraction
Extraction of phenolic compounds was performed, with slight
modifications, based on previous studies [25,28]. Blueberries (15 g)
were added to a solution containing acetone, water, and formic acid
(60:30:10 v/v) and homogenized using an Ultraturrax blender (IKA
Labortechnik, Staufen, Germany) at 20,300 rpm. The homogenate was
left at room temperature for 15 min and then centrifuged at 25°C at
4025 g for 5 min. The supernatant was collected and the pellet was
further washed with 25 ml of the extraction solvent; after further
centrifugation, the two supernatants were combined and the organic
solvent was removed by vacuum leaving an aqueous extract.
Hydrophilic and amphiphilic extracts of blueberry fruits for the
assessment of in vitro ferric reducing antioxidant potential (FRAP) and
total radical-trapping antioxidant parameter (TRAP) were obtained
according to the method of Pellegrini et al. [29]. Blueberry fruits were
pooled, mixed, and homogenized under nitrogen flux. A weighed
amount of sample (1 g) was extracted twice with 4 and 2 ml of water
in a shaking bath for 15 min at room temperature. The homogenates
were then centrifuged at 1000 g for 10 min and the supernatants
collected. The residue from pulp was extracted twice following the
same procedures using 4 and 2 ml volumes of acetone.
Analysis of blueberry phenolics
The phenolic profile of the blueberry extracts was assessed by an
RP-HPLC system equipped with a diode array detector [30]. The
column, an Inertsil ODS-3 column (4.6Ã?250 mm i.d., particle size
5 μm), was held at 40 °C and eluted at a flow rate of 0.5 ml min−1
. The
mobile phases consisted of acetonitrile (solvent A) and water:formic
acid (90:10 v/v) (solvent B); all solvents were HPLC grade. Elutionwas
performed following linear gradient steps: start condition 8% A in B,
then 11% A in 25min, 23% A in 22min, 45% A in 16min, 75% A in 3min,
75% A for 8 min, and 8% A in 5 min. Before injection (20 μl), samples
were diluted 1:10 (v/v) in mobile phase B and filtered through a 0.45
μmHAmembrane filter (Millipore Corp., Bedford, MA, USA).
Anthocyanin glycosides were identified on the basis of their absor-
bance spectra compared to published data [30]. All of the anthocyanin
monoglycosides were quantified as cyanidin 3-glucoside equivalents
[31]. Phenolic acids (chlorogenic acid, o-coumaric acid, ferulic acid,
gallic acid) and flavonols (quercetin 3-galactoside, quercetin 3-gluco-
side) were identified by comparing absorbance spectra and retention
times with those of authentic reference compounds.
Total antioxidant capacity (TAC) assays
Total antioxidant capacity in vitro (milk and blueberry extracts) and
in vivo (plasma samples) was assessed using the FRAP and the TRAP
assays, respectively, for the measurement of reducing and chain-
breaking antioxidant potential [32,33]. The FRAP assay is based on the
reduction of the Fe3+
?TPTZ complex to the ferrous form at low pH
monitored at 595 nmby a Sunrise absorbance plate reader (Tecan Italia
Srl, Segrate, Italy) [32].Briefly, 160 μl of FRAP reagent, prepared daily,
was mixed with 30 μl of water and 10 μl of diluted sample; the
absorbance at 595 nmwas recorded after a 30-min incubation at 37°C.
FRAP valueswere obtained by comparing the absorption changes in the
testmixturewith those obtained fromincreasing concentrations of Fe2+
and expressed as μmol of Fe2+
per liter of plasma or μmol g−1
The TRAPmethod is based on the protectionprovided by antioxidants
on the fluorescence decay of R-phycoerythrin (lag phase) during a
controlled peroxidation reaction [33].Inbrief,50 μl of diluted samplewas
added to 75 μlofPBS(pH7.4),15 μl of R-PE (fc 4.30Ã?10−3
μg μl
), and
60 μl of ABAP (fc 7.5 mM); the reaction kinetic at 38 °C was recorded for
60 min (λex =495 nm, λem=570 nm) by a Tecan GENios standard
fluorescence plate reader spectrometer (Tecan Italia Srl). The length of
the lag phase, automatically calculated, was used to assess TRAP values,
expressed as μmol g−1
or μmol L−1
for extracts and plasma, respectively.
In vitro study on milk addition to blueberry extracts
The hydro- and amphiphilic extracts of blueberries were used for
the in vitro study of milk association. Stock solutions of 50-fold-
diluted milk were prepared in MilliQ water. To 300 μl of blueberry
extract, 1000 μl of diluted milk and 700 μl of MilliQ water were added
[34]. After an incubation for 15 min at room temperature, the samples
were centrifuged for 15 min at 2683 g at 4°C and the supernatant was
collected. The pellet was redissolved in 500 μl of acetate buffer for
analysis. Both supernatant and pellet obtained from each hydro- and
amphiphilic extract were analyzed with the FRAP method as
described above.
In vivo study design
The acute ingestion model is a reliable tool to test the contribution
of antioxidant-rich food on the endogenous antioxidant defenses. The
Table 1
Nutritional composition of the milks used in the studya
Whole milk Semi-skimmed milk Skimmed milk
Energy content (kcal) 64 46 33
Proteins (g) 3.20 3.15 3.20
Carbohydrates (g) 4.80 4.85 4.90
Lipids (g) 3.60 1.55 0.10
Calcium (mg) 120 120 120
100 ml of product.
770 M. Serafini et al. / Free Radical Biology & Medicine 46 (2009) 769?774working window of the study is free from interference by different
variables (food intake, physical exercise, energy expenditure, etc.). The
study was approved by the ethics committee of the Istituto Nazionale
di Ricerca per gli Alimenti e la Nutrizione and all participants gave
their written consent. Eleven healthy volunteers (six men, five
women), nonsmokers, normolipidemic, taking no supplements or
medications, and free of any pathology, were recruited. Mean age and
BMI of the subjects were 31±5 years and 22.1±3 kg m−2
For 2 days before the intervention, the subjects followed a low-
antioxidant diet, avoiding phenolic-rich foods, namely, all fresh fruit
and vegetables and derived products (tea, coffee, fruit juices, wine,
chocolate). After an overnight fast, the volunteers were divided into
two groups and, following a crossover design, they consumed either
(a) 200 g of blueberry fruits plus 200 ml of water or (b) 200 g of
blueberry fruits plus 200 ml of whole milk. The intervention was
repeated after 1 week, this time swapping the treatments. Venous
samples at baseline (T0) and at 1 h (T1), 2 h (T2), and 5 h (T5)after
blueberry consumption were collected in EDTA- and heparin-
Vacutainers and centrifuged at 2500 rpm for 10 min at 4°C and the
plasma obtained was stored at −80°C for analysis.
Biochemical analyses
Plasma thiols were determined spectrophotometrically bymonitor-
ing the absorbance at 412 nm after addition of Ellman's reagent [35].
Uric acid, total cholesterol, and triacylglycerol concentrations in plasma
were quantified using colorimetric kits (GiesseDiagnostics,Rome, Italy).
Plasma vitamin C was immediately stabilized by adding metapho-
sphoric acid in the ratio of 1:1; ascorbic acid and dehydroascorbic acid
were determined by HPLC with electrochemical detection [36].
Determination of phenolic acids in plasma samples
Plasma phenolic acids were detected after enzyme and acid
hydrolysis of the conjugated forms as described by Azzini et al. [37].
Hydrolysis of conjugateswas performed by adding 100 μl ascorbic acid
(1%), 100 μl acetic acid (0.2 M), and 50 μl β-glucuronidase type HP2
from Helix pomatia (134300 U/ml) to 500 μl plasma. After incubation
at 37 °C for 2 h, proteins were precipitated with 700 μl methanol:HCl
(3 N) 1:1 v/v. Polyphenols were extracted with 2 ml ethyl acetate,
followed by stirring and sonication (2?3 min) before centrifugation at
1700 g for 5min. The extraction procedurewas repeated twice and the
ethyl acetate extracts were combined, reduced to dryness in vacuo,
and reconstituted in 200 μl of HPLCmobile-phasemethanol:H2O (1:1)
for analyses.
Quantitative analyses were performed with an ESA HPLC with a
coulometric electrode array detector [37]. A Supelcosil LC-18 column
(25 cmÃ?4.6 mm, 5 μm) was used, protected by a Perisorb Supelguard
LC-18 guard cartridge. Chromatography was performed at 30 °C with a
flow rate of 0.8 ml/min. This method employs a binary gradient:
solvent A, 0.02 M NaH2PO4?H2O adjusted to pH 2.8 with 85%
orthophosphoric acid; solvent B, methanol. It was programmed as
follows: 10% B at the start, increasing to 30% B over 7 min and to 33% B
over 28.5 min, increasing to 45% B over 19.5 min, held for 8.5min, and
reaching the final condition of 100% 24min later. The setting potentials
were 60,120, 200, 340, 480, 620, 760, and 900mV. Sample peaks were
identified by matching themwith standard peaks (caffeic acid, ferulic
acid; Sigma Chemical Co., St. Louis, MO, USA) on the basis of their
retention time and the accuracy ratio between adjacent channels.
The results obtained are expressed as means±standard deviation
of the mean (SDM) for the in vitro study, whereas values from the in
vivo study are expressed asmeans±standard error of themean (SEM).
The normal distribution of variables was confirmed by the Kolmo-
gorov?Smirnov test. A two-tailed Student t test was performed in
every sample, comparing the values of T0 with the various time points
postingestion as in previous studies [21]. The area under the curve
(AUC) of phenolic concentrations in plasma was calculated for the 5-h
period after each food intervention and the comparison of AUCs
between treatments was performed by t test. A p value of b0.05 was
considered significant. All statistical treatments were performed with
the Kaleida Graph version 4.0 program.
Phenolic composition and in vitro TAC of blueberry
The phenolic composition of blueberry fruits is described in Table
2. Five anthocyanins were identified: galactosides, arabinosides, and
Table 2
Phenolic composition of blueberry fruits
Phenolic compound mg 100 g−1a
Cyanidin galactoside 2.50
Cyanidin glucoside 0.93
Cyanidin arabinoside 1.74
Delphynidin galactoside 13.12
Delphynidin glucoside 1.11
Delphynidin arabinoside 8.70
Malvidin galactoside 26.16
Malvidin glucoside 1.60
Malvidin arabinoside 16.57
Petunidin galactoside 9.83
Petunidin glucoside 1.10
Petunidin arabinoside 5.95
Peonidin galactoside 2.02
Peonidin glucoside 0.99
Anthocyanins total 92.32
Chlorogenic acid 24.19
o-Cumaric acid 5.54
Ferulic acid 0.08
Gallic acid 1.00
Phenolic acids total 30.81
Quercetin galactoside 1.08
Quercetin glucoside 0.43
Quercetin xb
Flavonols total 2.01
Extracted in acetone/acidified water.
Glycosylated quercetin.
Fig. 1. In vitro total antioxidant capacity of blueberry extracts. Blueberry fruits were
pooled, mixed, and homogenized under nitrogen flux. A weighed amount of sample
(1 g) was extracted twice with 4 and 2 ml of water in a shaking bath for 15 min at room
temperature. The homogenates were then centrifuged at 1000 g for 10 min, and the
supernatants were collected (hydrophilic extract). The residue from pulp was extracted
twice following the same procedures using 4 ml and 2 ml volumes of acetone
(amphiphilic extract). Chain-breaking antioxidant potential (TRAP) and reducing power
(FRAP) were determined as described under Materials and methods. Results are
expressed as μmol g−1
of extract (mean±SDM; n=3).
771 M. Serafini et al. / Free Radical Biology & Medicine 46 (2009) 769?774glucosides of malvidin, delphinidin, cyanidin, petunidin, and peoni-
din. Anthocyanin derivatives represent 73.4% of the quantified
polyphenols, followed by phenolic acids with 24.5% (78.5% repre-
sented by chlorogenic acid) and flavonols with 2.1%.
The TAC values of blueberry fruits were 19.65±2.62 μmol g−1
extract in the TRAP assay and 15.75±0.35 μmol Fe+2
extract in the
FRAP assay (Fig. 1). The water-soluble fraction contributed a higher
percentage (69.7 and 60.3%, respectively, for TRAP and FRAP) than the
amphiphilic fraction (30.3% for TRAP and 39.7% for FRAP).
In vitro study on food association
The addition of various types ofmilk to blueberry extracts caused a
decrease in FRAP values of the supernatants, with whole milk
reducing more effectively than skimmed and semiskimmed milk
(Fig. 2). The extent of the antioxidant decrease was linearly associated
with the amount of lipids from the different milks (r=0.99901 for
hydrophilic extract; r=0.99659 for amphiphilic extract).
Effects of blueberry ingestion on markers of plasma antioxidant status
and lipid metabolism
Plasma concentrations of ascorbic acid, uric acid, thiols, total
cholesterol, and triacylglycerol did not change after blueberry
supplementation (Table 3). Plasma markers of TAC rose significantly
after blueberry ingestion (Fig. 3), reaching a maximum peak at 5 h,
+6.1%, pb0.001, and +11.1%, pb0.05, respectively, for FRAP and TRAP
assays. Plasma concentrations of caffeic acid and ferulic acid increased
significantly after blueberry ingestion at all time points. Different from
TAC, caffeic and ferulic acid reached Cmax (96±17 and 49±10 nM,
respectively) at 1 h postingestion and steadily decreased thereafter to
a final value of 42±7 nM for caffeic acid and 18±8 nM for ferulic acid
5 h postingestion (Fig. 3).
In vivo effects of milk addition on markers of plasma antioxidant status
and lipid metabolism
After the ingestion of blueberries alone, plasma concentrations of
vitamin C, uric acid, and thiol groups did not differ significantly from
the values obtained after consumption of milk and blueberries (Table
4). Total cholesterol concentrations increased significantly from
170±10 mg dl
up to a maximum value of 181±9 mg dl
at 5 h
postingestion (t test; pb0.001). Changes in plasma triacylglycerol
concentrations were not detected. The overall extent of absorption of
caffeic acid, but not of ferulic acid, as determined with estimates of
AUC, was reduced significantly by the addition of milk to the
blueberries (t test; p=0.039) (Fig. 4). Caffeic and ferulic acids appeared
in plasma 1 h after ingestion of the berries (Fig. 5), although milk
resulted in a marked reduction in their absorption. A significant
reduction of Cmax for caffeic acid (−49.7%, pb0.001) and ferulic acid
(−19.8%, pb0.05) was observed with respect to the levels reached
after ingestion of blueberries with water. The ingestion of blueberries
Fig. 2. Effect of milk addition on the in vitro reducing power (FRAP) of blueberry
extracts. Stock solutions of 50-fold-dilutedmilkwere prepared inMilliQwater. To 300 μl
of blueberry extract, 1000 μl of diluted milk and 700 μl of MilliQ water (control) were
added. After an incubation for 15 min at room temperature, samples were centrifuged
for 15 min at 2683 g at 4°C and the supernatant was collected. The pellet was
redissolved in 500 μl of acetate buffer for analysis. Both supernatant and pellet obtained
from each hydro- and amphiphilic extract were analyzed with FRAP as described under
Materials and methods. Results are expressed as means±SDM of three independent
experiments and represent the percentage of FRAP decrease in the supernatant of
blueberry extracts after milk addition relative to water (0% decrease). SM, skimmed
milk; SSM, semiskimmed milk; WM, whole milk.
Table 3
Effects of supplementation with 200 g of blueberry plus 200 ml of water on plasma
concentrations of single antioxidants and lipid markers in 11 healthy volunteers
Baseline 1 h 2 h 5 h
Ascorbic acid (mg dl
) 1.01±0.12 1.07±0.07 1.05±0.12 1.14±0.11
Uric acid (mg dl
) 4.30±0.40 4.39±0.50 4.43±0.50 4.39±0.50
Thiol groups (mg dl
) 612±25 598±18 605±22 629±47
Total cholesterol (mg dl
) 184±7 181±7 191±9 188±7
Triacylglycerols (mg dl
) 61.3±6.9 59.2±7.4 57.7±7.0 59.4±6.5
Fig. 3. In vivo effect of blueberry ingestion on markers of plasma antioxidant status and
phenolic bioavailability. Plasma levels of caffeic acid (□) and ferulic acid (○) and Δ
concentration relative to baseline values of plasma FRAP (■) and TRAP (●)after
ingestion of 200 g of blueberries plus 200 ml of water are shown. Experimental details
can be found under Materials and methods. Results are expressed as means±SEM
(n=11). Caffeic and ferulic acid plasma concentrations are significantly different from
baseline (paired t test; pb0.001 for all levels except for ferulic acid at 5 h, pb0.05).
pb0.001 for FRAP and pb0.05 for TRAP for time comparison at 5 h vs baseline.
Table 4
Effects of supplementation with 200 g of blueberries plus 200 ml of milk on plasma
concentrations of single antioxidants and lipid markers in 11 healthy volunteers
Baseline 1 h 2 h 5 h
Ascorbic acid (mg dl
) 1.08±0.10 1.19±0.10 1.10±0.12 1.19±0.09
Uric acid (mg dl
) 5.1±0.50 4.94±0.50 4.95±0.50 4.91±0.50
Thiol groups (mg dl
) 586±13 610±15 607±12 630±32
Total cholesterol (mg dl
) 170±10 176±10 178±11 181±9⁎
Triacylglycerols (mg dl
) 56.6±7.0 54.7±6.8 55.9±5.6 58.8±6.7
⁎ pb0.001 compared with baseline (paired t test).
772 M. Serafini et al. / Free Radical Biology & Medicine 46 (2009) 769?774and milk was not associated with any increase in TRAP and FRAP
plasma concentrations (Fig. 5). As after the ingestion of blueberries
with water, there was no time association between markers of
phenolic absorption and markers of antioxidant capacity (Fig. 5).
Effects of blueberry ingestion on markers of plasma antioxidant status
In agreement with previous studies [26,27], the results show that
blueberry ingestion results in a significant increase in (i) endogenous
plasma antioxidant defenses and (ii) caffeic and ferulic acid levels. The
presence of caffeic acid in plasma but not in the food matrix suggests
that postingestion metabolism of the blueberry phenolics results in
their appearance in the circulatory system. Caffeic acid might be
derived from the hydrolysis of chlorogenic acid (5-caffeoylquinic
acid). Previous intervention studies have reported that such a
transformation occurs after ingestion of artichoke [37] and coffee
[38] by human volunteers. As no 5-caffeoylquinic acid was detected in
the plasma, we hypothesize that 5-caffeoylquinic acid is hydrolyzed
and caffeic acid partly absorbed in the small intestine. The testedmeal
contained 48.38 mg of 5-caffeoylquinic acid, with an average intake of
0.74mg/kg body wt. The plasma Cmax of caffeic acid 1 h postconsump-
tion was 0.599 μg/kg body wt. If we assume that one molecule of 5-
caffeoylquinic acid produces one molecule of caffeic acid, the Cmax of
caffeic acid reached in plasma 1 h after ingestion of blueberry is
equivalent to 0.08% of 5-caffeoylquinic acid intake.
Despite the low bioavailability of phenolic acids and the lack of
changes in the levels of ascorbic acid, thiol groups, and uric acid, we
observed an increase in plasma antioxidant defenses. In both
treatments, there was no association between plasma concentrations
of caffeic and ferulic acids andmarkers of plasma TAC. After blueberry
ingestion, phenolics peaked 4 h before the increase in TAC, and when
milk was utilized, the increase in phenolics did not translate into any
positive changes in plasma TAC. Maximum increase of phenolic acids
after blueberry ingestion was in the midnanomolar range for caffeic
and ferulic acid (96 and 49 nM),whereas the increases in plasma TRAP
and FRAP were 101 and 49 μM, respectively. This 100-fold difference
makes it highly questionable whether these hydroxycinnamates
contribute to the endogenous redox network. In agreement with
previous evidence [39,40], our results highlight the lack of correlation
between plasma endogenous antioxidant defenses and circulating
levels of dietary phenolics. It is possible that synergistic interactions
between the redox component of the network and/or the presence of
other molecules such as phenolic metabolites might contribute to the
antioxidant effects of blueberry fruits in vivo.
In vivo and in vitro effects of milk addition
The in vitro addition of milk to blueberry extracts resulted in a
precipitation of blueberry antioxidants, leading to a decrease in TAC
values of the supernatants, with full-fat milk showing the highest
degree of inhibition. This finding is in agreement with the data from
Langley-Evans [41] that showed the greatest reduction in in vitro FRAP
values of black tea with the addition of whole milk (28% reduction)
compared to semiskimmed (22% reduction) and skimmed milk (12%
reduction). The data suggest a possible novel involvement of lipids in
the interaction between milk, protein, and antioxidant ingredients of
blueberry fruits. Polypeptide chains located around the fat globule
membrane in the milk increase with the increasing of lipid content of
the whole milk, supplying a more favorable environment for their
linkage with phenolics compared to lower fat milks [42]. However,
more in vitro experiments are needed to understand the role of lipids
in the inhibitory effect of milk on antioxidant activity.
In humans, in agreementwith previous evidence obtainedwith tea
and chocolate [15,19,21], we have shown that themilk reduces plasma
TAC and the Cmax of caffeic acid and ferulic acid, occurring after the
ingestion of blueberries. The plasma Cmax of caffeic acid 1 h after
blueberry consumption was 8.7 ng/ml. This is equivalent to 0.08% of
the 5-caffeoylquinic acid intake, and when the blueberry supplement
was ingested with milk this fell to 0.04% of intake.
The reduction of phenolics absorption in the presence of milk is in
agreement with Reddy et al. [43] and Serafini et al. [21], who reported
that the peak concentrations of plasma flavan-3-ols were significantly
lower after consumption of black tea with milk, milk chocolate, and
chocolate with milk. However, other studies have failed to show
significant differences in plasma concentrations of epicatechin [44],
total catechins [45], flavonols [18], and TAC [20,46] after ingestion of
chocolate and tea with milk.
Our results on TAC are in agreement with other studies in which
increases in plasma TAC after the ingestion of green tea [15], black tea
[15,19], and chocolate [21] are reduced by coconsumption of milk. The
increase in plasma TAC observed by Leenen et al. [20] with black tea
was +2%, which is very close to the 3% coefficient of variation in the
FRAP assay [32] and negligible compared to increases of +52 and +65%
Fig. 4. In vivo effect of milk addition on overall absorption of phenolics. The AUC of
plasma concentrations of caffeic acid and ferulic acid for the 0- to 5-h period after
consumption of 200 g of blueberries plus 200 ml of water or whole milk is shown.
Experimental details can be found under Materials and methods. Results are expressed
as means±SEM (n=11). ⁎pb0.05 for between-treatment comparison.
Fig. 5. In vivo effect of milk addition on markers of plasma antioxidant status and
phenolic bioavailability. Plasma levels of caffeic acid (□) and ferulic acid (○) and Δ
concentration relative to baseline values of plasma FRAP (■) and TRAP (●)after
ingestion of 200 g of blueberries plus 200 ml of whole milk are shown. Experimental
details can be found under Materials and methods. Results are expressed as means
±SEM (n=11). pb0.05 for FRAP for time comparison at 2 h vs baseline.
773 M. Serafini et al. / Free Radical Biology & Medicine 46 (2009) 769?774observed in previous studies [15,19], narrowing, in this way, the
observational window. In other studies in which chocolate was
utilized as the antioxidant source, Schroeter et al. [44] and Serafini
et al. [21] reached opposite conclusions, utilizing different proportions
of milk lipids: 3% in the former, with no effect of milk, and 30% in the
latter, with a clear inhibitory effect [47]. It is a matter of fact that the
discrepancy of the results in humans is remarkable, with half the
reports suggesting a lack of effect [18,20,43?46] and the other half
suggesting an inhibitory effect of milk [15,17,19,21].
The present study suggests, for the first time, that consumption of
a fruit rich in antioxidants, such as blueberry, in association with
whole milk decreases its ability to increase plasma endogenous
antioxidant defenses and to deliver into the circulation bioactive
molecules such as caffeic acid. More extensive research is needed to
clarify the role played by incorrect food association on the bioactivity
of dietary antioxidants in vivo.


ok I am not going to try reposting that.

Here's a pdf of the article



Thanks for posting.

Even though the study used milk, it mentions properties of protein in general that could interfere with phenolic absorption. I'll be eating blueberries on their own from now on...

Wonder if this has any implications towards Indigo-3G ?