Effect of tea on iron-absorption

Abstract

The effect of tea on iron absorption was studied in human volunteers. Absorption from solutions of FeCl3 and FeSO4, bread, a meal of rice with potato and onion soup, and uncooked haemoglobin was inhibited whether ascorbic acid was present or not. No inhibition was noted if the haemoglobin was cooked. The effect on the absorption of non-haem iron was ascribed to the formation of insoluble iron tannate complexes. Drinking tannin-containing beverages such as tea with meals may contribute to the pathogenesis of iron deficiency if the diet consists largely of vegetable foodstuffs.

Full text

Gut,
1975,
16,
193-200
The
effect
of
tea
on
iron
absorption
P.
B.
DISLER,
S.
R.
LYNCH,
R.
W.
CHARLTON,
J.
D.
TORRANCE,
T.
H.
BOTHWELL,
R.
B.
WALKER,
AND
FATIMA
MAYET
From
the
South
African
Medical
Research
Council
Iron
and
Red
Cell
Metabolism
Unit,
Departments
of
Medicine
and
Pharmacology,
University
of
the
Witwatersrand,
Johannesburg,
South
Africa,
the
Department
of
Botany,
University
of
Washington,
Seattle,
USA,
and
the
Department
of
Medicine,
University
of
Natal,
Durban,
South
Africa
SUMmARY
The
effect
of
tea
on
iron
absorption
was
studied
in
human
volunteers.
Absorption
from
solutions
of
FeCl3
and
FeSO4,
bread,
a
meal
of
rice
with
potato
and
onion
soup,
and
uncooked
haemoglobin
was
inhibited
whether
ascorbic
acid
was
present
or
not.
No
inhibition
was
noted
if
the
haemoglobin
was
cooked.
The
effect
on
the
absorption
of
non-haem
iron
was
ascribed
to
the
formation
of
insoluble
iron
tannate
complexes.
Drinking
tannin-containing
beverages
such
as
tea
with
meals
may
contribute
to
the
pathogenesis
of
iron
deficiency
if
the
diet
consists
largely
of
vegetable
foodstuffs.
Evidence
is
mounting
that
the
absorption
of
iron
from
individual
food
items
is
profoundly
affected
by
the
composition
of
the
meal
as
a
whole.
For
example,
egg
iron
is
very
poorly
absorbed,
but
the
percentage
is
considerably
increased
by
drinking
orange
juice
(Callender,
Marney,
and
Warner,
1970),
while
a
smaller
percentage
of
the
iron
in
soybean
is
absorbed
if
black
beans
form
part
of
the
meal
(Martinez-Torres
and
Layrisse,
1973).
Our
interest
in
tea
was
aroused
during
a
study
of
the
absorption
of
iron
from
maize
meal
porridge
served
with
sugar
containing
ferrous
sulphate
and
ascorbic
acid.
When
tea
was
drunk
with
the
meal
the
absorption
figures
seemed
lower
than
might
have
been
expected.
The
possibility
that
tea
might
inhibit
iron
absorption
did
not
appear
to
have
been
studied,
and,
since
it
is
a
popular
drink
in
a
number
of
countries
where
iron
deficiency
is
a
major
nutritional
problem,
it
was
decided
to
undertake
a
formal
investigation.
Materials
and
Methods
SUBJECTS
The
volunteers
who
took
part
in
the
present
investi-
gation
were
multiparous
Indian
housewives
living
in
a
municipal
housing
scheme
at
Chatsworth
near
Durban.
It
has
been
established
previously
that
iron
deficiency
is
common
among
such
individuals
(Mayet,
Adams,
Moodley,
Kleber,
and
Cooper,
1972).
Their
mean
age
was
40
years
(range
26-60
Received
for
publication
21
November
1974.
years).
Written
consent
was
obtained
after
the
nature
of
the
investigation
had
been
explained
to
them.
On
two
consecutive
mornings
after
an
overnight
fast
the
subjects
drank
a
solution
of
one
of
the
iron
compounds
or
ate
one
of
the
standard
meals.
The
iron
compound
or
food
iron
was
labelled
with
2-5
,uc
55Fe
on
the
one
morning
and
with
2-5
,uc
59Fe
on
the
other,
and
200
ml
of
either
warm
tap
water
or
tea
was
drunk
immediately
afterwards.
Nothing
more
was
eaten
or
drunk
for
the
next
four
hours.
Two
weeks
later
the
subjects
reassembled
after
again
fasting
overnight
and
a
specimen
of
blood
was
collected
for
measurement
of
the
55Fe
and
59Fe
content,
haemoglobin
concentration,
serum
iron
concentration,
unsaturated
iron-binding
capacity,
and
serum
ferritin
concentration.
They
then
drank
50
ml
tap
water
containing
1-7
,umol
ascorbic
acid
and
0'54
,umol
iron
as
FeSO4.7H20
labelled
with
2
5
,uc
59Fe.
No
further
food
or
drink
was
allowed
for
four
hours.
Measurement
of
the
59Fe
content
of
a
second blood
sample
collected
after
a
further
14
days
enabled
the
absorption
of
this
'reference
iron
salt'
to
be
calculated
by
difference,
and
provided
an
index
of
each
individual's
absorbing
capacity.
IRON
COMPOUNDS
Every
subject
received
0
54
,mol
iron
on
each
of
the
two
mornings.
In
different
experiments
the
iron
was
administered
as
FeCl3.
6H20,
as
FeSO4.
7H20
together
with
1-7
,umol
ascorbic
acid,
as
rabbit
haemoglobin,
or
as
crystallized
rabbit
haem.
The
193

194
P.
B.
Disler,
S.
R.
Lynch,
R.
W.
Charlton,
J.
D.
Torrance,
T.
H.
Bothwell,
R.
B.
Walker,
and
F.
Mayet
iron
salts
were
freshly
dissolved
in
50
ml
tap
water,
but
the
haemoglobin
and
the
haem
were
adminis-
tered
in
50
ml
preserved
tomato
juice
or
in
gravy.
Radioactive
haemoglobin
was
prepared
by
injecting
male
New
Zealand
white
rabbits
intramuscularly
with
200
pkc
of
either
59Fe
or
55Fe.
After
some
weeks
blood
was
obtained
from
a
marginal
ear
vein.
The
erythrocytes
were
separated
by
centrifugation
and
washed
three
times
with
sterile
isotonic
sodium
chloride
solution.
They
were
suspended
in
distilled
water,
frozen
to
-
20°C
and
thawed,
and
membranes
were
separated
from
the
haemoglobin
solution
by
centrifugation
at
2500
x
g.
Radioactive
haemoglobin
in
a
dosage
of
2-5
,kc
was
used
in
each
study.
It
was
mixed
with
either
unlabelled
haemoglobin
or
mince
to
provide
0
54
,umol
iron
per
person.
Haem
was
extracted
from
the
haemoglobin
by
the
method
of
Labbe
and
Nishida
(1962).
Radioactive
haem
was
mixed
with
unlabelled
haem
to
provide
0-5
Humol
iron
and
2
5
,uc
per
person.
IRON
IN
BREAD
Two
loaves
of
white
bread
were
baked,
using
70-80
%
extraction
flour.
Sufficient
59FeC13
or
55FeCl3
was
mixed
into
the
dough
together
with
the
yeast
so
as
to
provide
0
54
kmol
iron
and
2-5
[kc/100
g
bread,
the
quantity
consumed
by
each
subject.
The
bread
was
eaten
without
butter
or
jam.
IRON
IN
RICE
WITH
POTATO
AND
ONION
SOUP
Sufficient
rice
intrinsically
labelled
with
55Fe
by
hydroponic
culture
(Hussain,
Walker,
Layrisse,
Clark,
and
Finch,
1965;
Layrisse,
Cook,
Martinez,
Roche,
Kuhn,
Walker,
and
Finch,
1969)
was
mixed
with
carrier
rice
to
provide
2-5
,uc
55Fe
and
45
g
dry
rice
per
person.
The
rice
was
soaked
overnight
in
water
and
then
boiled
until
no
excess
water
re-
mained.
It
was
divided
into
equal
portions
by
weigh-
ing
and
eaten
with
the
potato
and
onion
soup.
Soup
for
10
subjects
was
prepared
by
frying
1
kg
peeled
potatoes
and
500
g
onions
in
2
table-spoons
of
sunflower
seed
oil.
After
adding
850
ml
water
and
5
g
curry
powder
it
was
brought
to
the
boil
and
allowed
to
simmer
for
30
minutes.
During
this
time
5-37
pmol
iron
as
FeSO4.7H20
labelled
with
25
[tc
59Fe
was
added
together
with
56-79
,umol
ascorbic
acid.
The
thick
soup
was
thoroughly
mixed
in
a
Waring
blender,
divided
into
equal
portions,
and
eaten
with
the
rice.
The
meal
thus
contained
55Fe
as
the
label
for
the
intrinsic
rice
and
iron
and
59Fe
as
the
label
for
the
added
FeSO4.
In
a
second
experiment
the
meal
of
rice
with
potato
and
onion
soup
was
prepared
in
the
same
way
except
that
no
intrinsically
labelled
rice
was
used;
on
the
one
morning
the
extrinsic
label
in
the
soup
was
59Fe
and
on
the
other
55Fe.
HAEMOGLOBIN
IRON
IN
MINCED
LAMB
Enough
minced
lamb
to
provide
a
total
of
0
54
pmol
iron
per
person
(including
that
present
in
the
hae-
moglobin
gravy)
was
fried
in
oil
and
divided
into
equal
portions
by
weighing.
Isotopically
labelled
rabbit
haemoglobin
solution
providing
2-5
,tc
per
individual
was
added
to
the
frying
pan
which
had
been
used
to
cook
the
mince,
and
simmered
for
15
minutes
to
make
a
gravy.
Equal
portions
were
poured
over
the
fried
mince
helpings.
TEA
A
commercial
brand
of
tea
widely
used
by
the
people
of
Chatsworth
(Pot
O'Gold,
O.K.
Bazaars
Ltd)
was
selected.
The
200
ml
drunk
by
each
individual
was
prepared
from
5
g
dry
tea.
In
some
studies
40
ml
pasteurized
cow's
milk
was
added.
When
the
effect
of
tea
with
milk
was
compared
with
that
of
tea
without
milk,
an
extra
40
ml
water
was
added
to
the
latter
to
make
the
volumes
the
same.
ISOTOPIC
AND
CHEMICAL
METHODS
Blood
samples
(10
ml)
and
aliquots
of
standard
iron
solutions
and
foods
were
prepared
for
differential
radioactive
counting
by
the
method
of
Katz,
Zoukis,
Hart,
and
Dern
(1964).
The
quantities
of
55Fe
and
59Fe
in
the
processed
samples
were
determined
by
means
of
a
liquid
scintillation
system
(Insta-Gel,
Packard
Instrument
Company,
Downers
Grove,
Illinois)
and
a
Packard
Tri-Carb
AAA
spectrometer
(model
3375),
which
automatically
adjusted
for
quenching.
The
counting
efficiency
was
2400
for
55Fe
and
42
%
for
59Fe
at
optimal
gain
and
window
settings.
The
59Fe
activity
in
the
4
ml
blood
samples
collected
immediately
before
the
'reference
iron
salt'
was
administered,
and
two
weeks
later
was
assessed
(against
suitable
standards)
by
means
of
a
Packard
Auto-Gamma
Tri-Carb
(model
3001)
spectrometer.
All
figures
for
percentage
absorption
were
calculated
on
the
assumption
that
100
%
of
the
absorbed
radio-
activity
was
present
in
the
haemoglobin
of
circu-
lating
red
cells,
and
that
the
blood
volume
of
each
subject
was
65
ml/kg.
We
calculated
that
if
the
whole
of
each
test
dose
had
been
retained,
the
total
radiation
dose
averaged
over
a
period
of
13
weeks
would
have
been
approximately
20%
of
the
permissible
whole
body
burden
for
continuous
exposure
in
the
case
of
59Fe
and
0-2
%
in
the
case
of
55Fe
(International
Commission
on
Radiological
Protection,
1960).
Serum
iron
concentrations
were
measured
by
a
modification
(Bothwell
and
Finch,
1962,
p.
18)
of
the
method
of
Bothwell
and
Mallett
(1955)
in
which
sulphonated
bathophenanthroline
was
used
as
the
colour
reagent.
The
unsaturated
iron-binding
capacity
was
determined
by
the
method
of
Herbert,
Gottlieb.
Lau,
Gevirtz,
Sharney,
and
Wasserman

195
The
effect
of
tea
on
iron
absorption
Haematological
Data
Percentage
Absorbed
Haemoglobin
Serum
Iron
UIBC
Percentage
Serum
Water
Tea
Reference
(gldl)
(Mmol/l)
(jimolil)
Saturation
Ferritin
Iron
Salt
Transferrin
(,ug/l)
FeCl,
9F9
6-27
93-62
6-3
-
3-7
2-3
25
6
13-6
19-69
60-50
24
5
77
6-9
2-5
8
5
10-9
8
41
77-33
9-8
8
7-5
3-6
23-5
13-9
16-11
68-20
19
1
41
7-7
4-8
5
8
12-8
17-90
60-14
22-9
15
9
5
2-0
24-3
15-5
14-50
82
52
14-9
16
21-0
12-2
33-7
11.0
6-27
92-90
6-3
2
23-1
12-3
65-3
11-8
13-25
71-42
15-6
39
25-9
7-8
42-3
119
14-68
74-82
16-4
8
51-7
7-9
42-4
14-0
13-07
72-14
15-3
33
60-0
6-2
31-6
Mean
12-5
13-02
75-36
151
26-4
21-7
6-2
30
3
(SD
±
19-7)
(SD
±
39)
(SD
±
17-3)
FeSO4
+
ascorbic
acid
14-1
22-73
48-69
31-8
7
4-0
4-0
13-7
14-50
62-83
18-8
2
5
9
4.9
14-1
9
85
47-44
17-1
2
9-8
3
9
14-4
13-78
57
46
19-3
88
24-3
4-2
12-9
20-94
57-82
26-6
92
33-4
8
5
12-9
11-64
81-45
12-5
74
37
0
12-4
95
6-09
77-51
6-1
21
403
23-1
14-1
17-54
47-79
22-6
23
42-4
12-6
14
8
18-08
67-48
21-1
8
52
7
23-1
12
5
9-49
89
50
9-6
5
59
0
15-7
Mean
13
3
14-46
63-80
17-0
32-0
30-9
11-2
(SD
+
19-3)
(SD
±
75)
Table
I
Effect
of
tea
on
absorption
of
iron
from
solutions
of
FeC!3
and
FeSO4
+
ascorbic
acid
Haematological
Data
Percentage
Absorbed
Haemoglobin
Serum
Iron
UIBC
Percentage
Serum
Tea
Tea
with
Milk
Reference
Salt
(gldl)
(Mumol/t)
(,Amol/)
Saturation
Ferritin
Transferrin
(lug/t)
12-4
22-55
78-94
22-2
26
2-5
4-3
28-5
12-4
18-26
75
90
19-4
34
3-3
3-1
35-1
12-4
15-75
55
85
22-0
23
4-3
10-8
456
12-1
14-86
59-61
20-0
22
5
2
9-3
51-3
12-4
20-94
85
20
19-7
39
15-7
16-1
78-9
12-1
17-36
73
39
19.1
26
16-5
8-1
46-4
11-8
9-31
83-59
10
0
<2
22-4
42-1
73.9
11-2
13-96
93-08
13-0
2
36-5
33-6
90
5
Mean
12-1
16-62
75
70
18-2
21-7
13-3
15-9
56
3
(SD
i
12-0)
(SD
±
14-3)
(SD
i
22-2)
Water
Milk
14-0
17-54
51-19
25-5
102
9-2
1-8
12-4
11-10
70-53
13-6
40
15-5
9-4
9
0
6-62
78-22
7-8
<2
30
0
27-9
12-1
7-88
89-68
8-1
<2
31-2 23-8
13-4
16-83 59-61
28-7
9
42-2
10-0
12-1
11-28
99-52
10-2
<2
42-7
33-8
9.4
6-98
89
50
7-2
3
59
7
21
2
11
2
22-73
72-14
24-6
13
60-4 21-4
10-8
8-77
100-78
8-0
<2
77-3
25-0
Mean
11-6
12-19
79-02
14-9
19-4
40-9
19-4
(SD
±
22
2)
(SD
:
10-2)
Table
II
Effects
of
tea
and
milk
on
absorption
of
iron
from
a
solution
of
FeSO4
+
ascorbic
acid

196
P.
B.
Disler,
S.
R.
Lynch,
R.
W.
Charlton,
J.
D.
Torrance,
T.
H.
Bothwell,
R.
B.
Walker,
and
F.
Mayet
Haematological
Data
Percentage
Absorbed
Haemoglobin
Serum
Iron
UIBC
Percentage
Serum
Water
Tea
Reference
Salt
(g/dl)
(Gmol/l)
(Amol/l)
Saturation
Ferritin
Transferrin
(Gg/l)
12-1
16-11
39-02
29-2
124
6-0
0-5
2-5
7-0
2-69
98-09
2-7
2
2-3
0-8
76-1
11-2
13-60
5245
20-6
18
10-3
1.1
26-2
13-6
7
70
51
37
13-0 18
10-4
1-6
25-2
12-1
16-65
59
97
21-7
11
11-6
2-7
35
2
14-0
20
05
51-55
28-0
4
13-4
3-8
60-2
10-6
6-98
99-88
6-5
7
14-2
6-6
50-3
14-0
11-81
73
03
13-9
62
15-3
8-9
10-9
Mean
11-8
11-95
65
67
16-9
30-8
10-4
3-3
35
8
(SD
4-4)
(SD
d
3-0)
(SD
250)
Table
III
Effect
of
tea
on
absorption
of
iron
in
bread
Haematological
Data
Percentage
Absorbed
Haemoglobin
Serum
Iron
UIBC
Percentage
Serum
Intrinsic
Extrinsic
Extrinsic
Reference
(gldi)
(O.mol/l)
(imol/l)
Saturation
Ferritin
Iron
Iron
Intrinsic
Iron
Salt
Transferrin
(Ag/l)
Meal
followed
by
water
11-4
12-17
89-32
12-0
33
5
3
5-1
0-96
37-2
12-2
20-23
68-38
22-9
8
6-7
7-2
1-07
22-5
6-8
4
30
102-57
4-1
2
10-3
9-0
0-87
33
5
11-8
12-17
86-46
12-4
11
11-5 11-7
1*02
45-4
10-6
6-98
95
94
6-8
4
13-5
16-3
1-20
21-4
10-0
11-46
100-60
10-3
13
16-0
21-9
1-36
43-6
11-2
7-34
103-28
6-6
2
23-5
21-6
0-92
53-0
Mean
10-6
10-66
92-36
10-4 10-4
12-4
13-3
1-05
36-6
(SD:
6-1)
(SD±
68)
(SD
±
0-17)
(SD
11-8)
Meal
followed
by
tea
14-2
20-94
49-76
29-6
100
0-5
0-6
1-20
2-9
14-8
21-48
58-71
26-8
104
0-5
0-6
1-20
3
9
14-4
13-78
6534
17-4
68
0-8
1-4
1-70
68-2
13-2
8-77
76-97
10-2
56
1-7
2-7
1-58
11-8
12-4
13-96
75-54
156
12
2-2
3-4
1-54
16-6
128
1486
7751
16-0
40
3-3
3-6
1-09
35-7
13-2
13-96
94-87
12-8
2
6-1
6-1
1-00
53-0
Mean
13-5
15-39
71-24
17-8
54-6
2-2
2-6
1-33
27-4
(SD
±
2-0)
(SD
±
2-0)
(SD
+
0-27)
(SD
±
25-5)
Table
IV
Absorption
of
intrinsic
and
extrinsic
(supplemental)
iron
from
rice
with
potato
and
onion
soup
containing
100
mg
ascorbic
acid
Haematological
Data
Percentage
Absorbed
Haemoglobin
Serum
Iron
UIBC
Percentage
Serum
Water
Tea
Reference
(gldl)
(Mmol/l)
(Omol/l)
Saturation
Ferritin
Iron
Salt
Transferrin
(,ug/l)
11-4
23-63
53-16
30-8
69
2-7
0-1
0-4
12-2
12-17
51-37
19
2
15
8-8
3-4
26-1
11-2
11-28
80-19
12-3
3
9.9
3-4
50-6
10-6
8-77
83-06
9-6
12
10-6
2-8
54-6
11-0
13-07
63-19
17-1
7
12-0
2-8
68-2
9-9
6-98
80-01
8-0
3
12-3
2-4
51-9
13-3
17-54
52-27
25-1
76
13-0
0-1
14-1
12-2
22-73
52-45
30-2
30
17-2
4-8
11-5
Mean
11-5
14-52
64-46
19-0
26-9
10-8
2-5
34-7
(SD
i
4-1)
(SD
i
1-6)
(SD
±
24-7)
Table
V
Effect
of
tea
on
absorption
of
iron
from
rice
with
potato
and
onion
soup
containing
100
mg
ascorbic
acid

The
effect
of
tea
on
iron
absorption
(1967).
The
iron
content
of
digested
samples
of
food
was
estimated
by
a
modification
(Bothwell
and
Finch,
1962,
p.
26)
of
the
method
of
Lorber
(1927).
The
serum
ferritin
concentrations
were
measured
by
radioimmunoassay
using
the
method
of
Miles,
Lipschitz,
Bieber,
and
Cook
(1974).
Results
EFFECT
OF
TEA
ON
THE
ABSORPTION
OF
IRON
FROM
SOLUTIONS
OF
IRON
SALTS
The
drinking
of
tea
without
milk
was
found
to
inhibit
the
absorption
of
iron
from
a
solution
of
FeCls
(t
=
2-68,
p
=
<
0-05),
and
also
from
a
solution
of
FeSO4
containing
ascorbic
acid
(t
=
4-46,
p
=
<0-01)
(table
I).
Tea
with
milk
produced
much
the
same
effect
on
the
absorption
of
iron
from
a
solution
of
FeSO4
with
ascorbic
acid
(t
=
8-65,
p
=
<0-001)
as
did
tea
without
milk
(t
=
9-28,
p
=
<
0-001),
the
degree
of
inhibition
being
revealed
by
comparison
with
the
'reference
absorption'
figures
(table
II).
When
200
ml
milk
without
tea
was
drunk
after
the
solution
of
FeSO4
and
ascorbic
acid
absorption
was
also
inhibited
(t
=
3-44,
P
=
<0-01).
EFFECT
OF
TEA
ON
THE
ABSORPTION
OF
NON-
HAEM
FOOD
IRON
Tea
inhibited
the
absorption
of
iron
from
bread
(t
=
7-50,
p
=
<
0-001)
(table
Ill).
The
absorptions
of
the
intrinsically
labelled
rice
iron
and
the
extrinsic
iron
in
the
potato
and
onion
soup
were
closely
similar
whether
tea
was
drunk
with
the
meal
or
not
(table
IV).
This
observation,
together
with
the
con-
siderable
evidence
from
other
studies
that
all
the
non-haem
iron
in
a
meal
forms
a
common
pool
within
the
lumen
of
the
gut
(Cook,
Layrisse,
Martinez-Torres,
Walker,
Monsen,
and
Finch,
1972;
Bjorn-Rasmussen,
Hallberg,
and
Walker,
1972;
Sayers,
Lynch,
Jacobs,
Charlton,
Bothwell,
Walker,
and
Mayet,
1973;
Sayers,
Lynch,
Charlton,
Bothwell,
Walker,
and
Mayet,
1974),
indicated
that
an
extrinsic
label
could
be
used
to
assess
the
effect
of
tea
on
the
absorption
of
iron
in
the
meal.
Accordingly,
eight
further
subjects
consumed
the
rice
with
potato
and
onion
soup
on
successive
mornings,
the
extrinsic
label
being
59Fe
on
the
one
occasion
and
55Fe
on
the
other.
When
tea
was
drunk
with
the
meal
there
was
a
significant
inhibition
of
absorption
(t
=
6-74,
p
=
<0-001)
(table
V)
in
spite
of
the
presence
of
100
mg
ascorbic
acid.
Haematological
Data
Percentage
Absorbed
Haemoglobin
Serum
Iron
UIBC
Percentage
Serum
Water
Tea
Reference
(gld!)
(Mmol/t)
(Mmol/l)
Saturation
Ferritin
Iron
Salt
Transferrin
(isg/f)
Uncooked
haemoglobin
10-4
13-96
77-87
15-2
102
5-5
1-7
2-1
14-6
18-08
72-67
20-0
25
11-5
2-0
12-2
13-6
22-73
71-60
24-1
14
11-3
3-1
25-2
12-8
7-70
60-50
11-3
13
4-7
3-5
18-1
14-0
23-63
61-40
27-7
5
5-6
3-7
29-2
12-0
26-85
47-79
36-0
33
8-0
4-1
13-3
12-8
31-68
59-07
34-8
5
22-6
4-4
26-9
13-6
19-33
46-54
29-3
32
13-0
5-6
8-7
15-8
11-28
81-45
12-2
3
13-9
7-5
37-6
12-8
15-75
91-65
14-6
14
11-4
8-5
96-6
7-5
3-94
89-50
4-2
<2
23-4
10-2
74-3
8-6
6-09
75-18
7-5
<2
45-3
17-7
83-7
Mean
12-4
16-75
69-60
19-7
12-8
14-7
6-0
35-7
(SD
+
11-4)
(SD
4-
5)
(SD
i
31-6)
Cooked
haemoglobtn
13-2
17-54
46-18
27-5
54
2-6
3-0 2-8
11-8
15-93
46-18
25-6
32
16-7
7-5
15-2
12-8
12-35
54-77
18-4
29
10-3
7-7
12-9
13-6
19-51
36-70
34-7
31
10-0
8-1
28-2
12-4
17-36
46-36
27-2
5
11-2
11-2
48-6
11-4
12-71
57-28
18-2
<2
15-6
11-2
25-3
12-4
13-25
61-93
17-6
10
25-1
13-7
87-3
13-6
13-96
48-51
22-3
29
10-9
24-1
55-5
14-0
13-78
50-30
21-5
59
24-1
27-8
1-9
12-1
11-46
65-16
15-0
21
8-2
28-9
27-2
Mean
12-7
14-79
51-34
22-8
27-1
13-5
14-3
30-5
(SD
±
7-0)
(SD
+
9-2)
(SD
±
26-6)
Table
VI
Effect
of
tea
on
absorption
of
haemoglobin
iron
197

198
P.
B.
Disler,
S.
R.
Lynch,
R.
W.
Charlton,
J.
D.
Torrance,
T.
H.
Bothwell,
R.
B.
Walker,
and
F.
Mayet
Haematological
Data
Percentage
Absorbed
Haemoglobin
Serum
Iron
UIBC
Percentage
Serum
Cooked
Uncooked
Reference
(gldi)
(jsmol/l)
(itmol/i)
Saturation
Ferritin
Haemoglobin
Haemoglobin
Iron
Salt
Transferrin
(yg/l)
14-1
31-50
22-20
58
7
5
3-8
0
9
45-3
14-6
3544
37-77
40
4
21
8-1
4-7
22-7
15-2
42-24
4099
519
21
10-5
7-8
22-3
14-4
20-23
37
95
34-8
18
11-3
4
5
15-0
14-1
16-47
40-10
29-1
25
14-6
15-2
18-3
8-2
7-34
77-87
8-6
20
15-3
5-2
28-3
14-8
31-15
5549
36-0
2
16-0
12-5
65-2
10-6
9-85
77-87
11-2
14
16-2
4-7
28-7
13-7
15-57
5495
22-1
118
18-2
9-9
3-5
12
4
19-33
54-06
26-3
7
23-4
15-6
32-5
Mean
13-2
22-91
49-93
31-9
25
0
13-7
8-1
28-2
(SD
+
55)
(SD
±
50)
(SD
±
17-1)
Table
VII
Absorption
of
iron
from
cooked
and
uncooked
haemoglobin
ingested
with
tea
EFFECT
OF
TEA
ON
THE
ABSORPTION
OF
HAEM
IRON
Tea
significantly
inhibited
the
absorption
of
hae-
moglobin
iron
from
a
solution
of
uncooked
rabbit
haemoglobin
in
tomato
juice
kt
=
3-89,
P
=
<0
005)
(table
VI).
Of
greater
practical
importance,
however,
was
the
finding
that
tea
had
no
significant
effect
on
the
absorption
of
haemoglobin
iron
from
the
fried
lamb
mince
with
rabbit
haemoglobin
gravy
(t
=
0-28,
p
=
>
0-70).
The
conclusion
that
tea
did
not
affect
the
absorption
of
haemoglobin
iron
from
cooked
food
was
checked
by
a
comparison
between
the
absorption
of
cooked
and
uncooked
rabbit
haem-
oglobin
iron
administered
in
tomato
juice
and
followed
by
a
cup
of
tea
on
each
occasion
(table
VII).
The
absorption
of
iron
from
the
uncooked
haem-
oglobin
was
significantly
less
than
from
the
cooked
haemoglobin
(t
=
4-66,
p
=
<0005).
In
different
experiments
the
absorption
of
haemoglobin
iron
without
tea
was
very
similar
whether
the
haemoglobin
was
uncooked
or
cooked
(table
VII),
and
this
is
in
agreement
with
previous
reports
(Callender,
Mallett,
and
Smith,
1957;
Turnbull,
Cleton,
and
Finch,
1962).
It
therefore
seemed
justifiable
to
conclude
that
tea
inhibits
the
absorption
of
haemoglobin
iron
only
if
it
has
not
been
cooked.
Finally,
the
effect
of
tea
on
the
absorption
of
crystallized
rabbit
haem
was
examined.
No
inhibition
was
found,
the
mean
figures
(±SD)
in
nine
subjects
being
11V8%
(±3
5)
with
water
and
10-6
%
(±
5
0)
with
tea.
RELATIONSHIP
BETWEEN
SERUM
FERRITIN
CON-
CENTRATION
AND
IRON
ABSORPTION
Only
where
ferrous
sulphate
and
ascorbic
acid
had
been
administered
were
there
enough
absorption
results
to
permit
the
rate
of
iron
absorption
to
be
correlated
with
the
serum
ferritin
concentration.
The
percentage
iron
absorptions
from
the
different
E
6
10
20
30
40
50
60
70
80
90
Iron
Absorption
(%1
Fig
The
relationship
between
the
serum
ferritin
concentration
(mean
+
2SE)
and
the
absorption
of
iron
from
a
solution
of
FeSO4
with
ascorbic
acid.
The
iron
solution
was
followed
by
a
drink
of
water
(A)
or
tea
(B).
experiments
were
grouped
into
decades
and
were
plotted
against
the
mean
log
serum
ferritin
con-
centrations
(see
fig).
A
straight
line
relationship
was
revealed
such
that
y
=
-
0-01
8x
+
1
650
(r
=
-
0-67,
p
=
<0-001).
When
tea
was
drunk
with
the
iron
solution
the
slope
was
steeper
(y
=
-
0029x
+
1F488)
(r
=
-0
50,
p
=
<0-01).
The
difference
between
the
regression
coefficients
of
the
two
lines
was
statistically
significant
(t
=
2-26,
p
=
<
0
05).
Discussion
The
results
of
the
present
study
indicate
that
tea
inhibits
the
absorption
of
non-haem
iron
to
a
significant
extent.
The
effect
was
seen
with
a
solution
of
ferric
chloride,
with
a
solution
of
ferrous
sulphate

The
effect
of
tea
on
iron
absorption
199
plus
ascorbic
acid,
and
with
the
iron
in
bread
and
in
a
rice
meal,
and
was
similar
whether
the
tea
contained
milk
or
not.
No
attempt
was
made
to
investigate
the
mechanism
responsible
for
the
interference
with
absorption,
but
it
seems
likely
that
it
was
due
to
the
tannins
in
the
tea.
Tannins
form
coloured
complexes
with
ferric
iron
(Finar,
1956),
and
a
blackish
dis-
coloration
was
seen
when
0
54
pmol
iron
as
FeCI3
was
added
to
200
ml
tea.
Much
the
same
effect
was
produced
by
the
quantity
of
ferrous
sulphate
with
ascorbic
acid
used
in
the
present
study.
The
formation
of
iron
complexes
within
the
intestinal
lumen
may
profoundly
affect
iron
absorption
(Conrad,
1970).
Such
complexes
may
be
soluble
or
insoluble.
Some
of
the
agents
which
form
soluble
complexes
facilitate
iron
absorption,
eg,
ascorbic
acid,
while
others
such
as
bicarbonate
(Benjamin,
Cortell,
and
Conrad,
1967)
and
EDTA
(Brise
and
Hallberg,
1962)
reduce
the
availability
of
the
iron.
If
the
complex
is
insoluble
it
is
not
absorbed.
The
extent
to
which
iron
can
be
absorbed
from
food
is
probably
largely
dependent
on
the
relative
concentrations
of
various
com-
plexing
agents
present
in
the
meal.
It
follows
that
supplemental
iron
salts
are
subject
to
the
same
influences
as
the
intrinsic
iron
present
in
individual
foodstuffs,
and
it
has
repeatedly
been
shown
that
they
are
absorbed
to
a
similar
degree
(Cook,
Layrisse,
Martinez-Torres,
Walker,
Monsen,
and
Finch,
1972;
Bjorn-Rasmussen
and
Hallberg,
1972;
Sayers,
Lynch,
Jacobs,
Charlton,
Bothwell,
Walker,
and
Mayet,
1973;
Sayers,
Lynch,
Charlton,
Bothwell,
Walker,
and
Mayet,
1974).
Further
evidence
of
this
was
obtained
in
the
present
study
(table
IV).
Haem
iron
is
absorbed
as
such,
and
it
is
only
within
the
mucosal
epithelial
cells
that
the
iron
is
liberated
from
the
porphyrin
(Weintraub,
Weinstein,
Huser,
and
Rafal,
1968).
Luminal
chelators including
ascorbic
acid
have
been
shown
not
to
influence
its
absorption
(Callender,
Mallet,
and
Smith,
1957;
Turnbull,
Cleton,
and
Finch,
1962;
Conrad,
Benjamin,
Williams,
and
Foy,
1967),
and
the
sig-
nificant
inhibition
of
the
absorption
of
uncooked
haemoglobin
iron
by
tea
was
therefore
surprising.
An
agent
capable
of
chelating
ionic
iron
is
most
unlikely
also
to
be
able
to
form
complexes
with
haem
(Conrad,
1970),
and
indeed
the
absorption
of
crystallized
haem
was
found
not
to
be
inhibited.
The
possibility
that
the
inhibition
of
haemoglobin
iron
absorption
by
tea
might
be
due
to
an
effect
on
the
globin
was
therefore
considered.
The
tanning
of
leather
is
thought
to
involve
the
formation
of
cross-
links
between
collagen
fibres,
the
phenolic
groups
of
the
vegetable
tannins
probably
attaching
to
the
peptide
bonds
between
the
amino
acids
by
hydrogen
bonding
(Haslem,
1966).
Possibly
the
uncooked
globin
was
'tanned'
by
the
tea,
and
thereby
rendered
less
susceptible
to
hydrolysis
by
the
proteolytic
enzymes
of
the
digestive
juices.
If
this
occurred
then
less
haem
would
be
released
and
less
would
be
available
for
absorption.
The
observation
that
tea
had
no
inhibitory
effect
if
the
haemoglobin
were
cooked
strengthened
this
possibility,
since
cooking
denatures
the
globin
but
does
not
affect
the
absorp-
tion
of
haemoglobin
iron
(Callender
et
al,
1957;
Turnbull
et
al,
1962).
Since
meat
is
almost
invariably
cooked
before
it
is
eaten,
little
nutritional
significance
can
be
attached
to
the
inhibition
of
the
absorption
of
uncooked
haemoglobin.
In
those
communities
where
meat
is
an
important
dietary
constituent
the
iron
nutrition
is
generally
satisfactory,
but
iron
deficiency
is
rife
when
the
average
diet
of
the
population
consists
very
largely
of
vegetable
staples.
Iron
is
poorly
absorbed
from
wheat,
maize,
or
rice
meals
(Martinez-
Torres
and
Layrisse,
1973;
Sayers
et
al,
1973;
Sayers
et
al,
1974),
and
tea
may
aggravate
the
nutritional
problem.
Tea
is
drunk
during
meals
by
many
South
Africans
of
Indian
extraction,
and
this
is
also
the
custom
in
other
parts
of
the
world.
Since
preliminary
observations
(unpublished)
indicate
that
coffee
has
a
similar
effect,
the
implications
may
be
even
wider.
An
interesting
peripheral
observation
was
the
inverse
correlation
between
the
serum
ferritin
concentration
and
the
percentage
absorption
of
iron
(see
fig).
The
relationship
was
similar
to
that
found
by
Cook,
Lipschitz,
Miles,
and
Finch
(1974),
and
confirms
the
value
of
the
serum
ferritin
concentration
as
a
measure
of
the
body's
need
for
iron.
The
steeper
slope
of
the
line
when
tea
was
drunk
can
be
ascribed
to
the
effective
sequestration
of
a
proportion
of
the
iron
in
unabsorbable
tannin
complexes.
The
pre-
valence
of
iron
deficiency
among
the
group
as
a
whole
was
underscored
by
the
finding
that
the
serum
ferritin
concentration
was
below
the
lower
limit
of
normality
of
10
,ug/l
in
a
third
of
them
(Jacobs,
Miller,
Worwood,
Beamish,
and
Wardrop,
1972;
Lipschitz,
Cook,
and
Finch,
1974).
This
work
was
supported
in
part
by
grants
from
the
International
Atomic
Energy
Agency,
Vienna,
the
Atomic
Energy
Board,
South
Africa,
and
the
South
African
Sugar
Association.
The
authors
are
grateful
to
Mrs
Shirley
Lichtigfeld,
Miss
Fawzia
Khan,
Miss
Premilla
Maharaj,
and
Mrs
Felicity
Hurwitz
for
their
indispensable
technical
assistance.
References
Benjamin,
B.
I.,
Cortell,
S.,
and
Conrad,
M.
E.
(1967).
Bicarbonate-
induced
iron
complexes
and
iron
absorption:
one
effect
of
pancreatic
secretions.
Gastroenterology,
53,
389-396.
Bjorn-Rasmussen,
E.,
and
Hallberg,
L,
and
Walker,
R.
B.
(1972)

200
P.
B.
Disler,
S.
R.
Lynch,
R.
W.
Charlton,
J.
D.
Torrance,
T.
H.
Bothwell,
R.
B.
Walker,
and
F.
Mayet
Food
iron
absorption
in
man.
I.
Isotopic
exchange
between
food
iron
and
inorganic
iron
salt
added
to
food;
studies
on
maize,
wheat
and
eggs.
Amer.
J.
clin.
Nutr.,
25,
317-323.
Bothwell,
T.
H.,
and
Finch,
C.
A.
(1962).
Iron
Metabolism,
Little,
Brown,
Boston.
Bothwell,
T.
H.,
and
Mallett,
B.
(1955).
The
determination
of
iron
in
plasma
or
serum.
Biochem.
J.,
59,
599-602.
Brise,
H.,
and
Hallberg,
L.
(1962).
Iron
absorption
studies.
II.
Acta
med.
scand.,
171,
suppl.
376,
7-73.
Callender,
S.
T.,
Mallett,
B.
J.,
and
Smith,
M.
D.
(1957).
Absorption
of
haemoglobin-iron.
Brit.
J.
Haemat.,
3,
186-192.
Callender,
S.
T.,
Marney,
S.
R.,
and
Warner,
G.
T.
(1970).
Eggs
and
iron
absorption.
Brit.
J.
Haemat.,
19,
657-665.
Conrad,
M.
E.
(1970).
Factors
affecting
iron
absorption.
In
Iron
Deficiency.
Pathogenesis:
Clinical
Aspects:
Therapy,
edited
by
L.
Hallberg,
H.
G.
Harworth,
and
A.
Vannotti,
pp.
87-120.
Academic
Press,
New
York
and
London.
Conrad,
M.
E.,
Benjamin,
B.
I.,
Williams,
H.
L.,
and
Foy,
A.
L.
(1967).
Human
absorption
of
hemoglobin
iron.
Gastroenterology,
53,
5-10.
Cook,
J.
D.,
Layrisse,
M.,
Martinez-Torres,
C.,
Walker,
R.,
Monsen,
E.,
and
Finch,
C.
A.
(1972).
Food
iron
absorption
measured
by
an
extrinsic
tag.
J.
clin.
Invest.,
51,
805-815.
Cook,
J.
D.,
Lipschitz,
D.
A.,
Miles,
L.
E.
M.,
and
Finch,
C.
A.
(1974).
Serum
ferritin
as
a
measure
of
iron
stores
in
normal
subjects.
Amer.
J.
clin.
Nutr.,
27,
681-687.
Finar,
I.
L.
(1956).
Organic
Chemistry.
Volume
2.
Stereochemistry
and
the
Chemistry
of
Natural
Products,
p.
599.
Longmans,
Green,
London,
New
York,
Toronto.
Haslem,
E.
(1966).
Chemistry
of
Vegetable
Tannins,
pp.
7-9.
Academic
Press,
New
York
and
London.
Herbert,
V.,
Gottlieb,
C.
W.,
Lau,
K.
S,
Gevirtz,
N.
R.,
Sharney,
L.,
and
Wasserman,
L.
R.
(1967).
Coated
charcoal
assay
of
plasma
iron
binding
capacity
and
iron
using
radio
isotope
dilution
and
hemoglobin-coated
charcoal.
J.
nucl.
Med.,
8529.541.
Hussain,
R.,
Walker,
R.
B.,
Layrisse,
M.,
Clark,
P.,
and
Finch,
C.
A.
(1965).
Nutritive
value
of
food
iron.
Amer.
J.
clin.
Nutr.,
16,
464-471.
International
Commission
on
Radiological
Protection
(1960).
Report
of
Committee
II
on
Permissible
Dose
of
Internal
Radiation,
1959
(I.C.R.P.
Publication
No.
2),
pp.
85-89.
Pergamon
Press,
Oxford,
for
I.C.R.P.,
London.
Jacobs,
A.,
Miller,
F.,
Worwood,
M.,
Beamish,
M.
R.,
and
Wardrop,
C.
A.
(1972).
Ferritin
in
the
serum
of
normal
subjects
and
patients
with
iron
deficiency
and
iron
overload.
Brit.
med.
J.,
4,
206-208.
Labbe,
R.
F.,
and
Nishida,
G.
(1957).
A
new
method
of
hemin
iso-
lation.
Biochim.
biophys.
Acta
(Amst.),
26,
437.
Layrisse,
M.,
Cook,
J.
D.,
Martinez,
C.,
Roche,
M.,
Kuhn,
I.
N.,
Walker,
R.
B.,
and
Finch,
C.
A.
(1969).
Food
iron
absorp-
tion:
a
comparison
of
vegetable
and
animal
foods.
Blood,
33,
430-443.
Lipschitz,
D.
A.,
Cook,
J.
A.,
and
Finch,
C.
A.
(1974).
A
clinical
evaluation
of
serum
ferritin
as
an
index
of
iron
stores.
New
Engl.
J.
Med.,
290,1213-1216.
Lorber,
L.
(1927).
Einfache
Mikrokolorimetrische
Eisenbestimmungs
methode.
Biochem.
Z.,
181,
391-394.
Martinez-Torres,
C.,
and
Layrisse,
M.
(1973).
Nutritional
factors
in
iron
deficiency:
food
iron
absorption.
Clin.
Haemat.,
2,
339-352.
Mayet,
F.
G.
H.,
Adams,
E.
B.,
Moodley,
T.,
Kleber,
E.
E.,
and
Cooper,
S.
K.
(1972).
Dietary
iron
and
anaemia
in
an
Indian
community
in
Natal.
S.
Afr.
med.
J.,
46,
1427-1430.
Miles,
L. E.
M.,
Lipschitz,
D.
A.,
Bieber,
C.
P.,
and
Cook,
J.
D.
(1974).
Measurement
of
serum
ferritin
by
a
2-site
immuno-radiometric
assay.
Analyt.
Biochem.,
in
press.
Sayers,
M.
H.,
Lynch,
S.
R.,
Charlton,
R.
W.,
Bothwell,
T.
H.,
Walker,
R.
B.,
and
Mayet,
F.
(1974).
Iron
absorption
from
rice
meals
cooked
with
fortified
salt
containing
ferrous
sulphate
and
ascorbic
acid.
Brit.
J.
Nutr.,
31,
367-375.
Sayers,
M.
H.,
Lynch,
S.
R.,
Jacobs,
P.,
Charlton,
R.
W.,
Bothwe
11
T.
H.,
Walker,
R.
B.,
and
Mayet,
F.
(1973).
The
effects
of
ascorbic
acid
supplementation
on
the
absorption
of
iron
in
maize,
wheat
and
soya.
Brit.
J.
Haemat.,
24,
209-218.
Turnbull,
A.,
Cleton,
F.,
and
Finch,
C.
A.
(1962).
Iron
absorption.
IV.
The
absorption
of
hemoglobin
iron.
J.
clin.
Invest.,
41,
1897-1907.
Weintraub,
L.
R.,
Weinstein,
M.
B.,
Huser,
H.
J.,
and
Rafal,
S.
(1968).
Absorption
of
hemoglobin
iron:
the
role
of
a
heme-splitting
substance
in
the
intestinal
mucosa.
J.
clin.
Invest.,
47,
531-539.