1. food and drink

Food consumption
patterns
in seven countries13
Daan Kromhout,
Ancel Keys,
Annemarie
Jansen,
Alessandro
Bozidar
S Simic,
and Hironori
in the 1 960s
Christ Aravanis,
Ratko Buzina,
Menotti,
Srecko Nedeljkovic,
Toshima
ABSTRACT
At the
gate the relations
Finland,
Greece,
between diet and cardiovascular
Italy, Japan, The Netherlands,
end
of the
1950s
the
Flaminio
Fidanza,
Maija Pekkarinen,
Seven
Countries
Study
Simona
was
Giampaoli,
designed
to investi-
1960s food consumption
data were collected
the record method.
In Finland
the intake ofmilk,
potatoes,
very high. A similar
but lower intake pattern
was observed
and pastry consumption
was high in Italy; and
was high in the United
edible fats, and sugar
in The Netherlands.
States; cereal and alcoholic
products
was
Fruit, meat,
drink consumption
high in Yugoslavians
except for those in Belgrade.
In Greece the intake ofolive
oil and fruit was high and the Japanese
cohorts
were characterized
by a high consumption
of fish, rice, and soy products.
These
differences
in food consumption
patterns
have lessened during the past 25 y.
Am J Clin Nuir 1989;49:889-94.
KEY WORDS
bread
Dietary
consumption
survey,
food consumption
Introduction
At the end ofthe
1950s the Seven Countries
Study was
designed
to investigate
relations
between
diet and cardiovascular
diseases
(1-3). Sixteen
cohorts
were selected
in
Finland,
Greece,
Italy, Japan,
The Netherlands,
United
States,
and Yugoslavia.
The base-line
survey
was fobbowed-up
after 5 and 10 y by repeat surveys.
Since then
only mortality
data were collected,
which are now almost
complete
for a subsequent
25-y period.
During
the base-line
survey
13 000 men, aged 4059 y, were medically
examined.
Information
on diet was
patterns,
record
method
in Ilomantsi
close to the Russian
border; the other was in West
Finland
in Poytya and Mellila.
The dietary
survey took place
in September,
1959. Both in East and West Finland
30 men
took part. The food consumption
of each man was weighed
on seven consecutive
days by university
students
majoring
in
nutrition
and was recorded
as the edible parts ofthe raw products.
Italy had three cohorts
participating
in the study: inhabitants
of Montegiorgio
t
a country
From the Department
village
in the
ofEpidemiobogy,
region
of Marche
National
Institute
on
of Pub-
‘‘
collected
in random
samples
from
each cohort
5). Detailed
data
on
by use
of
the record
method
(4,
food consumption
patterns
have been published
only for 9 of the
16 cohorts
(4, 6). Therefore,
the food intake
data were
coded once again into a standardized
form by one person. Then the foods were summarized
in a limited
number of food groups.
The average
daily consumption
per
person
of these food groups
was calculated
for each cohort. The differences
in food consumption
pattern
between the 16 cohorts
are described.
lic Health
Perugia,
dietary
surveys
in the different
cohorts
1989;49:889-94.
Printed
in USA.
© 1989 American
The Netherlands;
of Health,
Diseases,
and Nutrition,
ofEpidemiology
Italy; the Laboratory
Medical
Metabolic
Rome,
School,
Italy;
the Internal
Belgrade,
Zagreb,
University of
and Biostatis-
Yugoslavia;
Clinic
B,
the De-
grade, Yugoslavia;
Department
oflnternal
Supported
Medicine,
by a grant
from
Kurume,
the Queen
Japan.
Wibhelmina
Fund,
Amster-
The Netherlands.
Address reprint requests
to D Kromhout,
Department
of EpidemiNational
Institute
of Public Health and Environmental
Protection, P0 Box 1, 3720 BA Bilthoven,
The Netherlands.
Received December
18, 1987.
obogy,
The Finn participation
in the Seven Countries
Study consisted of two cohorts. One cohort was situated in East Finland
Am J C/in Nutr
Bilthoven,
University
of Helsinki,
Helsinki,
Finland;
the
Medical
Faculty,
University
of Belgrade,
Beland Kurume University, School ofMedicine,
Third
dam,
ofihe
Perugia,
tics, National
Institute
University
ofBebgrade,
partment
of Nutrition,
Institute
of Hygiene,
3
Description
Protection,
stitute for Diabetes,
Endocrinology
and
Yugoslavia;
the Institute
ofFood
Sciences
2
Methods
and Environmental
the Division of Epidemiology,
School of Public Health, University
of
Minnesota,
Minneapolis,
MN; the Greek Society for the Study of Atherosclerosis
and the Medical Center ofAthens,
Athens, Greece; the In-
Society
Accepted
for Clinical
for publication
Nutrition
May 3 1,
1988.
889
Downloaded from ajcn.nutrition.org by guest on August 28, 2014
diseases. Sixteen cohorts were selected in
United States, and Yugoslavia.
During the
from random
samples
of these cohorts
by use of
KROMHOUT
890
the east coast; inhabitants
to the city of Bologna;
Montegiorgio
the food
1960. A total of 36 men
son’s
data
were
not
ofCrevalcore,
a country
village close
and railroad
personnel
in Rome.
In
intake
data were collected
in April,
completed
the 7-d recording.
One per-
used
for
analyses
because
only
4 d were
recorded.
In Crevalcore
the dietary
survey was completed
the months
ofJune
and July, 1960. Twenty-nine
men took
in this
survey
Crevalcore
products.
and
there
were
all the foods
The
food
no dropouts.
were recorded
intake
data
from
in
part
and
In Montegiorgio
as the edible
parts
Rome
collected
were
of raw
in
July, 1969. Fifty men were asked to record their food consumption for 7 d. Of 49 participants
the records
were suitable
for
processing.
The foods were recorded
in their prepared
state.
Yugoslavia
had five cohorts
participating
in the study. Two
cohorts
were from
sisted ofinhabitants
the Croatian
ofsix places
part of Yugoslavia
on the Dalmatian
and concoast in the
consisted
of farm laborers
from the country
village
of Vebika
Krsna,
60 km southeast
of Belgrade;
factory
workers
from a
small place called Zrenjanin,
60 km north ofBelgrade;
and Belgrade
university
professors.
The first round
of the dietary
survey in Velika
Krsna
was carried
out in October,
1962. The data
ofonly
five people
for 6 d were still available.
For this reason
it
was decided to process
4 1 men were requested
seven
consecutive
the food intake data ofMay,
1963 when
to record their food consumption
for
days.
The
data
of 22 of the 4 1 men
available.
One had to be disqualified
cause he had only recorded
his food
Zrenjanin
44 men were asked to take
in September,
1963. The data offour
fled because
they had only recorded
for 6 d. The dietary
survey in Belgrade
the months
of January
and February,
sors
were
asked
to record
their
food
were
their
food
consumption
was first carried
1964. Forty-one
consumption
it appeared
information
that
out in
profes-
for 7 d. The
data ofall participants
were able to be processed.
The
take data from these three cohorts
were processed
in a
manner.
All foods were recorded
as edible parts of
product.
The Netherlands
participated
in the study with one
In 1960, 5 1 men from Zutphen
took part in the dietary
by use of the 7-d record
method.
From the coding of
consumption
data
supplied
insufficient
still
from the processing
beconsumption
for 6 d. In
part in the dietary survey
people had to be disquali-
food inuniform
the raw
cohort.
survey
the food
five ofthe
participants
over the 7 d and one
had
person
had failed to separate
the hot meal into potatoes,
vegetables,
and meat. The data of these six people were not processed
for
the statistical
analyses.
All foods were weighed
and recorded
directly
before
consumption.
In the United
State
study.
of Minnesota
This survey
States,
group ofpeople
examined
were studied
respectively
from October
to December,
1962.
A total
employees
cooperated
was carried
of 30 men
of the US Railroad
in the
in the dietary
survey
of the
out during
1960-62.
The total
was made up ofthree
subgroups
that
from September
to October,
1960;
196 1; and from January
to March,
weighed
and
recorded
their
food
involved
were
Agies,
Paraskies,
Thrapsano,
and
Kastelli.
In
Corfu the villages
were Ano Korakiana,
Skriperon,
and San
Marco.
About
30 men were involved
in each dietary
survey.
However,
the original
7-d records
were no longer available.
It
was therefore
decided
to reconstruct
the diets of these cohorts
on the basis of results of the dietary
surveys
mentioned
in a
publication
by Keys et al (7). When no information
about the
consumption
available
food
of certain
foods,
eg, fruits
and vegetables,
balance
sheet data from Greece
in 1961-65
was
were
used as a substitute
(8).
Japan participated
in the study with two cohorts.
The first
cohort
was taken from Tanushimaru,
a village in the interior
ofJapan.
The food consumption
of24 men in this cohort was
recorded
for 7 d in November
and December,
1964 by dieticians
and
assistant
doctors
who
visited
the
families
at meal-
times. Snacks between
meals were recorded
by the women.
The
other cohort was situated
in Ushibuka,
a small fishing village
on the coast. In May and June, 1971 the food consumption
of
eight men was recorded
for 4 d. The method
was the same as
in Tanushimaru.
The original
7- and 4-d food intake data provided the starting
point for processing.
Nearly all the products
were given as edible parts ofraw foods. One exception
was rice,
which
was always
Processing
recorded
ofthefood
Fourteen
in cooked
intake
ofthe
data
16 dietary
and 1964 (Table
form.
surveys
1). Two surveys
took
place
between
were held around
1959
1970. The
7-cl record
method
was used in 14 of the 16 cohorts.
The 1-d
record
was used in the US railroad
and the 4-d record
in the
Japanese
Ushibuka
study.
ate and drank
during
sumed
were weighed
The participants
recorded
the period
ofresearch.
as well. In 1 3 of the
were recorded
as the edible part of the raw products
of the 16 as prepared
products.
This has consequences
following
foods:
cereal
products,
what
The amounts
16 surveys
the
legumes,
meat,
fish,
they
confoods
and in 3
for the
and
vege-
tables. Cereal products
and legumes
increase
in weight during
preparation.
Meat, fish, and vegetables
lose weight during preparation.
To compare
the different
cohorts,
all the quantities
of
these prepared
products
were converted
in a standardized
way
into
quantities
ofraw
products.
After data entry a check was made for incorrect
coding numhers together
with a frequency
distribution
of the quantities
concerned
of each
food.
From
this
it was
possible
to discover
extreme
personal
values. These values were printed
out according
to the
identity
number
and from this the original
food record could be checked
to see ifthe extreme
values were correct.
After checking,
the foods were put into a limited
number
of
food groups
consisting
of foods
with a similar
water
content
and nutrient
composition.
The sugar products
group consisted
of foods containing
at least 60% sugar, eg, jams and honey but
not chocolate
and soft drinks.
The alcoholic
drinks,
eg, beer,
wine, and liquor were converted
into grams ofpure
alcohol due
to their different
alcohol content.
All foods consumed
by the men in the different
cohorts
could be classified
in 15 homogeneous
food groups.
The 16th
food group was very heterogeneous
and included
products
such
as herbs, spices, sauces,
precooked
ready-to-eat
meals, yeast,
nuts,
chocolate,
soft drinks,
ice cream,
sumption
ofthese
16 food groups
day for each cohort.
etc. The average
was calculated
per person
conper
con-
sumption
for 1 d. The products
were recorded
in their prepared
state.
In Greece five dietary surveys were carried out between
1960
and 1965; three in Crete and two in Corfu. In Crete the villages
Results
The number
offoods
consumed
Yugoslavian
cohort
Vebika Krsna
varied from 52 in the
to 186 in the Dutch
Downloaded from ajcn.nutrition.org by guest on August 28, 2014
Makarska
area and inhabitants
ofa small place called Dalj near
Osijek on the Hungarian
border in the province
ofSlavonia.
In
April, 1960 24 men in the province
of Slavonia
recorded
their
food consumption
for seven consecutive
days as did 24 men in
the province
Dalmatia
in May, 1960. The foods were recorded
as edible parts ofthe raw products.
The other three Yugoslavian
cohorts
were from Serbia and
ET AL
FOOD
TABLE
CONSUMPTION
IN SEVEN
COUNTRIES
891
I
Summary
ofthe
collected
food consumption
Cohort
Country
data
per cohort
Year
Numberof
Record
persons
method
Month
Number
foods
of
consumed
d
Finland
1959
September
30
7
91
West*
Finland
1959
September
30
7
96
Montegiorgio*
Crevalcore*
Italy
Italy
Rome railroadt
Italy
Dalmatia*
Sbavonia*
Vebika Krsna*
Zrenjanin*
Yugoslavia
Yugoslavia
Yugoslavia
Yugoslavia
Yugoslavia
The Netherlands
1960
1960
1969
1960
1960
1963
1963
1964
1960
April
June-July
July
April
May
May
September
January-February
July-August
35
29
49
24
24
21
40
41
45
7
7
7
7
7
7
7
7
7
11
163
182
67
64
52
1 12
174
186
1960-62
September-March
30
1
147
1960-65
196 1-63
1964
197 1
February-September
March-September
November-December
May-June
31
37
24
8
7
7
7
4
174
126
Bebgrade*
Zutphent
US railroadt
United
Crete*
Corfu*
Tanushimaru*
Ushibuka*
Greece
Greece
Japan
Japan
Food recorded
*
States
as raw products.
t Food recorded
:1:Unknown,
cohort
as prepared
original dietary
Zutphen
(Table
products.
records are no longer
1). The
men
available.
in Vebika
Krsna
had
between
a very simple
dietary
pattern.
They ate a great deal of
bread and cheese but little ifany fruit, sugar, pastry, and
fish. The reason
why the Japanese
cohorts
scored
high
on the number
of foods consumed
was mainly
because
ofthe large variety offish and seaweed,
which amounted
to 27% ofthe total food consumption.
In Italy and Yugoslavia
there
was
a marked
difference
in food
consump-
In Finland,
tion between
those in the cities and the country
villages.
The choice ofproducts
was far greater in the cities.
From Table 2 it appears
that most cohorts
consumed
TABLE
2
Average
amount
ofvegetabbe
Cohort
foods
consumed
Country
per person
Bread
240
and
450
g bread/day.
The
Americans
and
Japanese,
however,
ate far bess bread. The quantity
in Japan was < 10 g/d. On the other hand the consumption
of cereal products
(mainly
rice) was very high in Japan,
almost
500 g/d. The consumption
of cereal products
in
Italy was also high but here it was mainly pastas that were
eaten.
The
Netherlands,
and
Dabmatia
(Yugosla-
via) the potato
consumption
was high (200-300
g/d).
Few potatoes
were consumed
in the Italian
cohorts,
in
Velika Krsna (Yugoslavia),
and in Ushibuka
(Japan).
Ja-
per day per cohort
Cereals
Potatoes
Legumes
Vegetables
108
104
194
140
260
200
198
Fruit
g
East
Finland
380
71
273
1
West
Finland
356
99
296
8
171
149
113
64
120
29
83
50
56
30
29
21
Montegiorgio
Italy
Crevalcore
Romerailroad
Dabmatia
Italy
Italy
Yugoslavia
Slavonia
Yugoslavia
Velika Krsna
Zrenjanin
Belgrade
Yugoslavia
Yugoslavia
Yugoslavia
358
349
249
435
440
794
425
244
Zutphen
The Netherlands
252
17
USraibroad
UnitedStates
97
26
Crete
Greece
380
30
Corfu
Greece
450
45
Tanushimaru
Japan
5
Ushibuka
Japan
10
5
6
2 14
7
129
28
136
86
40
40
34
28
191
150
6
1
22
115
1
15
12
245
179
185
145
252
2
227
124
1
171
233
190
30
191
464
150
30
191
462
497
95
103
174
26
449
34
79
222
42
82
Downloaded from ajcn.nutrition.org by guest on August 28, 2014
East*
892
KROMHOUT
TABLE 3
Average amounts
ofanimal
foods consumed
Cohort
per person
Country
ET
AL
per day per cohort
Meat
Fish
Eggs
Milk
Cheese
g
West
Finland
Finland
105
107
58
7
11
35
Montegiorgio
Crevalcore
Rome railroad
Dalmatia
Slavonia
Velika Krsna
Italy
Italy
Italy
Yugoslavia
Yugoslavia
Yugoslavia
85
154
226
1 17
188
35
39
9
22
54
24
33
77
4
434
Zrenjanin
Zutphen
USrailroad
Crete
Yugoslavia
Yugoslavia
The Netherlands
UnitedStates
Greece
Corfu
Greece
Tanushimaru
Ushibuka
Japan
Japan
East
Belgrade
pan was the only country
with a high consumption
of
legumes,
‘---90 g/d. This is because
soy products
were
classified
as legumes.
Vegetable
consumption
varied between
100 and 200 g/d among
most cohorts.
More than
200 g of vegetables
per day were consumed
by the railroad workers
in Rome,
in Ushibuka
(Japan),
Zutphen
(The Netherlands),
and Zrenjanin
(Yugoslavia).
The
fruit consumption
varied greatly from cohort
to cohort.
Very little fruit (< 10 g/d) was eaten by the Yugoslavian
cohorts
in Dabmatia,
Slavonia,
and Velika Krsna.
A lot
of fruit (460 g/d) was eaten by the men in Greece.
Here
it should be noted that the dietary
survey in Finland
was
carried
out in the berry season,
therefore,
fruit consumption was relatively
high at that time.
Little meat was eaten in Japan and Greece
(< 35 g/d)
(Table
3). A lot of meat (> 200 g/d) was eaten by the
Americans,
in Zrenjanin
(Yugoslavia),
and in Rome (Italy). No fish was eaten in Vebika Krsna (Yugoslavia).
In
contrast,
fish consumption
in Dalmatia
(Yugoslavia)
and Tanushimaru
(Japan)
was high (“-90 g/d) and very
high (--200
g/d) in Ushibuka
(Japan).
Fish consumption
varied between
10 and 60 g/d among
most ofthe cohorts.
Very few eggs were used in Corfu
(Greece)
and East
Finland
(5 and 1 1 g/d, respectively).
The most eggs, averaging l/d, were used in Slavonia
(Yugoslavia)
and Crevalcore (Italy). The rest of the cohorts
used an average
of
halfan
egg per day. No cheese was eaten in Japan. A large
amount
of cheese (200 g/d) was eaten in Velika
Krsna
(Yugoslavia).
Cheese
consumption
varied
between
10
and 35 g/d in most cohorts.
The variation
in milk consumption
was enormous.
Practically
no milk was drunk
in Japan
and
in Montegiorgio
(Italy);
the
milk
consump-
tion in Finland
was > 1 L/d.
In Japan very little edible fat was used (< 10 g/d) (Table 4). In the other cohorts
the use of edible fats varied
between
25 and 100 g/d. In Finland,
The Netherlands,
1192
1090
8
313
30
25
96
31
35
51
21
228
0
7
19
37
19
27
203
16
51
12
27
3 1
3
18
60
93
207
40
25
5
19
39
18
13
14
0
0
191
184
335
447
231
235
70
28
23
and the United
States mostly
saturated
fatty acids were
used and in the southern
European
countries
mostly unsaturated
fatty acids; mainly
olive oil in Greece and Italy
and sunflower
and olive oil in Yugoslavia
were used. The
consumption
of sugar products
was negligible
in Velika
Krsna
(Yugoslavia)
but
was
high
in Finland
and
The
Netherlands.
Hardly
any pastries
were eaten in Yugoslavia, Greece,
and Japan,
whereas
the consumption
of
pastries
in the United
States was extremely
high (nearly
lOOg/d).
There was considerable
variation
between
the cohorts
in the consumption
ofalcohol.
The lowest consumption
was in East Finland
(1 g/d) and the highest
in Dalmatia
(Yugoslavia)
(95 g/d). In Italy the consumption
was also
high (“70 g/d) and as in Dalmatia
the kind of alcoholic
drink was mainly wine. In Japan mainly strong alcoholic
drinks (sake and shochu)
were consumed.
The last category is a very heterogeneous
rest group.
On account
of
the large variety in foods it is impossible
to make a comparison
between
the
cohorts.
Discussion
In the past the dietary
patterns
ofthe
16 cohorts
of the
Seven Countries
Study were primarily
described
in terms
ofthe fatty acid composition
ofthe diet (9). Detailed
data
on the food consumption
pattern
ofalb cohorts
have not
been
published.
Therefore,
the food consumption
data
of all cohorts
were coded in a standardized
way by one
dietitian
in close collaboration
with the colleagues
in the
different
countries.
The earlier published
data (4, 6) on
the food consumption
pattern
of some cohorts
differ
sometimes
slightly from the data presented
in this paper.
However,
these
differences
do not influence
earlier
drawn
conclusions
about the characteristics
of the food
Downloaded from ajcn.nutrition.org by guest on August 28, 2014
70
212
175
138
273
35
35
8
8
19
18
FOOD
TABLE 4
Average amounts
ofremaining
Cohort
foods
consumed
Country
CONSUMPTION
per person perday
IN
SEVEN
893
COUNTRIES
per cohort
Edible fats
Pastries
Sugar products
Rest
100% Alcohol
g
East
Finland
96
West
Finland
Italy
Italy
Italy
72
63
58
51
Yugoslavia
Yugoslavia
88
70
Montegiorgio
Crevalcore
Romeraibroad
Dalmatia
Slavonia
Veika
Krsna
91
101
13
12
1
2
39
28
7
4
64
19
38
6
79
25
19
39
22
12
0
0
65
95
21
56
3
1
1
0
12
2
2
12
12
24
28
Yugoslavia
55
26
Belgrade
Yugoslavia
49
54
6
5
Zutphen
US railroad
Crete
The Netherlands
United States
79
33
72
29
3
29
24
95
6
91
Greece
95
20
0
15
107
Corfu
Greece
75
13
0
31
109
Tanushimaru
Ushibuka
Japan
3
13
0
18
23
Japan
7
26
5
25
25
consumption
pattern
of the different
cohorts
of the
Seven Countries
Study.
Each country
and even each region had, in the 1960s,
its own characteristic
food consumption
pattern.
In Finland the intake
of milk, potatoes,
edible fats, and sugar
products
was very high. A similar
but lower intake pattern was observed
in The Netherlands.
Meat, vegetable,
and fruit consumption
was high in the United
States,
Rome (Italy),
Belgrade,
and Zrenjanin
(Yugoslavia).
In
Italy the consumption
of cereals
and alcoholic
drinks
were high. The Dalmatia
(Yugoslavia)
cohort
was characterized
by a high alcohol
and fish intake.
In Velika
Krsna
(Yugoslavia)
the consumption
of bread
and
cheese was high. In Slavonia
(Yugoslavia)
cereal
products and egg consumption
were high. The Greek cohorts
were characterized
by a high intake ofolive
oil and fruit.
In Japan
a lot of fish, rice and, soy products
were consumed.
These
characteristic
food consumption
patterns
have
greatly
changed
during
the
last 25 y. Unfortunately,
with
the exception
of the Dutch
cohort,
the two Italian
cohorts ofCrevalcore
and Montegiorgio,
the two Greek cohorts Crete and Corfu, and the two Finn cohorts
the food
consumption
studies
were not carried
out at regular
times
by using the same method.
From
the Zutphen
study
it appears
that
the
consumption
of meat,
fruit,
al-
coholic
drinks,
and pastries
has increased
during the last
25 y whereas
consumption
ofbread,
potatoes,
milk products, and edible fats has decreased
considerably
(Kromhout D, unpublished
observation).
In the two Italian cohorts the consumption
of fruit has increased
during
the
last 20 y and that of cereals and legumes
has decreased
(10, 1 1). In Montegiorgio
an increase
was observed
in the
intake of milk, cheese, and meat (1 1). In the two Greek
cohorts
olive oil intake decreased
during the last 20 y and
the intake of alcohol
increased
(12). In the two Finn co-
horts the consumption
of meat and fruit increased
between
1959 and 1969 whereas
consumption
of sugar
products
decreased
(1 3). These changes
continued
after
1969. In addition,
the consumption
of bread and potatoes decreased
until about
1980. From
1969 onward
an
increase
in the consumption
of alcoholic
beverages
was
observed
(M Pekkarinen,
personal
communication,
1987).
Of the remaining
countries
no data from food consumption
studies are available.
However,
there are food
balance
sheet data available
for these countries
during
the period
196 1-77 (8). In the United
States the consumption
of vegetables,
fruit, meat,
edible
fats, sugar
products,
and alcoholic
drinks
has increased
and the
consumption
ofmilk
products
and eggs has decreased.
In
Yugoslavia,
the consumption
of vegetables,
fruit, meat,
eggs, edible
fats, milk products,
and sugar products
has
increased.
The
notable
changes
in the Japanese
food
pat-
tern are an increased
consumption
of fruit, meat, eggs,
milk products,
edible fats, and alcoholic
drinks and a decrease in cereals.
Because
of these changes
the differences
in food consumption
patterns
among
the cohorts
have
become
fewer. This does not mean that characteristic
differences
no longer
exist. It is important
for epidemiologic
research
to know
whether
the relative
position
of the cohorts
in the distribution
of different
foods was maintamed.
Because
of the back of follow-up
data this question could not be studied
in the 16 cohorts
of the Seven
Countries
Study. However,
for the seven countries
food
balance
sheet, data are available
for the periods
1961-65
and 1975-77
(8). The high correlation
coefficients
between the food balance
sheet data of 1961-65
and 197577 of the
seven
countries
tion of the countries
foods was maintained
indicate
that
in the distribution
(Table
5). These
the
relative
posi-
of the different
results
suggest
Downloaded from ajcn.nutrition.org by guest on August 28, 2014
Yugoslavia
Zrenjanin
KROMHOUT
894
TABLES
Pearson
RHMM
correlation
77 in seven
coefficients
countries
between
on the
basis
foods
offood
in 1961-65
balance
sheet
077*
Potatoes
0.68
Legumes
0.88t
Vegetables
0.89
Fruit
Meat
Fish
Eggs
Milk and cheese
0.98t
0.98t
0.95t
0.81*
I .OOt
Sugar
0.95t
products
the food
of the
consumption
l960s
can
data
be used
logic studies into the relation
from chronic
diseases
among
seven countries.
This paper
describes
the
but not the energy
the seven countries.
calculated
because
for each cohort.
ing the average
lected recently.
trients
will be
these chemical
the
in the
We thank
Japanese
collected
at the begin-
for prospective
epidemio-
between
diet and mortality
the different
cohorts
in the
food
consumption
pattern
and nutrient
intake
of the cohorts
in
Energy
and nutrient
intake
were not
local
food tables
were not available
In addition,
food composites
food intake ofeach
cohort have
In these food composites
energy
determined
chemically.
The
analyses
will become
available
foreseeable
future.
Keys A, Aravanis
C, Blackburn
HW, et al. Epidemiobogical
studies
related to coronary
heart disease: characteristics
of men aged 4059 in seven countries.
Acta Med Scand b967;460(suppb):l-392.
2. Keys
A. Coronary
heart disease
in seven countries.
Circulation
1970;4l(suppl
b):l-2ll.
3. Keys A. Seven countries:
a multivariate
analysis ofdeath
and coronary heart disease.
Cambridge,
MA: Harvard
University
Press,
1980.
4. Den Hartog C, Buzina
R, Fidanza
F, Keys A, Rome P. Dietary
studies and epidemiology
of heart diseases.
The Hague: Stichting
tot wetenschappelijke
voorlichting
op voedingsgebied,
1968.
5. Keys A, Kimura
N. Diet of middle-aged
farmers
in Japan. Am J
Clin Nutr b970;23:212-23.
6. Fidanza
F, Findanza
Alberti A. Rilevamento
dci consumi
alimentari di alcune
famiglie
in tre zone agriole
d’Italia.
Quaderni
Nutr
197 l;3 1: 139-88.
7. Keys A, Aravanis
C, Sdrin H. The diets ofmiddle-aged
men in two
rural areas ofGreece.
Voeding
1966; 27:575-86.
8. Food
and Agriculture
Organization
of the United
Nations.
Food
balance sheets and per caput food supplies.
Rome: Food and Agriculture Organization
ofthe United Nations,
1980.
9. Keys A. Coronary
heart disease in seven countries.
Circulation
b970;4b(suppl
1): 162-83.
representbeen coband nuresults
of
for publi13
Mr Ri Schlemper
for his translation
and help in coding
food intake
data. Mr C de Lezenne
Coulander,
Mr
10.
Fidanza
F, Fidanza
two rural
Metab
population
1975;
Abberti
groups
A. Food
of Italy
and
nutrient
followed
consumption
for ten years.
of
Nutr
18:176-89.
1 1. Fidanza
F, Fidanza
Alberti A, Cobi R, Mencarini
Contini
A. Food
and nutrient
consumption
of two rural Italian population
groups
followed for twenty years. Int J Vitam Nutr Res 1983;53:91-7.
12. Aravanis
C, boannidis
PJ. Nutritional
factors and cardiovascular
diseases in the Greek
Islands
Heart
Study.
In: Lovenberg
W, Yamori Y, eds. Nutritional
prevention
ofcardiovascubar
disease. New
York: Academic
Press, 1984:125-3S.
13. Pekkarinen
M. Dietary surveys in connection
with coronary
heart
disease studies in Finland.
In: New trends in nutrition,
lipid research and cardiovascular
diseases.
New York: AR Liss Inc. 198 1:
243-61.
Downloaded from ajcn.nutrition.org by guest on August 28, 2014
0.001.
ning
played a beading part in
1.
0.96t
0.75*
drinks
van Steen, and Mr BPM Bboemberg
the processing
and statistical
analyses
ofthe collected
data. Further,
we
thank Professor
L R#{228}s#{228}nen
and the assistants
ofProfessors
F Fidanza,
A Menotti,
and H Toshima
for their cooperation
in the translation
work. We are very grateful to Mrs B Christodoubou
for her help in reconstructing
the diets ofthe Greek cohorts.
References
0.01.
that
cation
1975-
r
Cereals
Ediblefats
Alcoholic
p
and
data
Foods
t p
ET AL