European Journal of Clinical Nutrition (2002) 56, 629–637
ß 2002 Nature Publishing Group All rights reserved 0954–3007/02 $25.00
www.nature.com/ejcn
ORIGINAL COMMUNICATION
Serum lipid profiles in Japanese women and men
during consumption of walnuts
M Iwamoto1, K Imaizumi1*, M Sato1, Y Hirooka2, K Sakai2, A Takeshita2 and M Kono1
1
Laboratory of Nutrition Chemistry, Division of Bioresource and Bioenvironmental Sciences, Graduate School, Kyushu University,
Fukuoka, Japan; and 2Research Institute of Angiocardiology and Cardiovascular Clinic, Kyushu University School of Medicine,
Fukuoka, Japan
Objective: To determine the serum cholesterol, apolipoproteins and LDL oxidizability in young Japanese women and men
during walnut consumption and to evaluate its active principle.
Design: Experimental study with a randomized design.
Subjects: Twenty healthy women and 20 healthy men.
Interventions: Subjects were randomly assigned to consume each of two mixed natural diets for 4 weeks in a cross-over design.
Reference and walnut diets were designed and the walnut diet had 12.5% of the energy derived from walnuts (44 – 58 g=day).
Results: The total cholesterol and serum apolipoprotein B concentrations, and the ratio of LDL cholesterol to HDL cholesterol
was significantly lowered in women and men when fed on the walnut diet, than when on the reference diet (P 0.05). The LDL
cholesterol concentration was significantly lowered in women on the walnut diet (0.22 mmol=l, P ¼ 0.0008), whereas this
decrease was not significant in men (0.18 mmol=l, P ¼ 0.078). The most prominent change in the fatty acid composition of the
cholesteryl esters from serum after the walnut diet was an elevation of a-linolenic acid in women (76%, P < 0.001) and men
(107%, P < 0.001). This elevation was negatively correlated to the change in LDL cholesterol in women (r ¼ 0.496, P ¼ 0.019)
and men (r ¼ 0.326, P ¼ 0.138). The LDL oxidizability in women was not influenced by the diets (P ¼ 0.19).
Conclusions: a-Linolenic acid in the walnut diet appears to be responsible for the lowering of LDL cholesterol in women.
Sponsorship: Kyushu University (Fukuoka, Japan) and the California Walnut Commission (California, USA).
European Journal of Clinical Nutrition (2002) 56, 629 – 637. doi:10.1038=sj.ejcn.1601400
Keywords: cholesterol; Japanese diet; a-linolenic acid; oxidized LDL; walnuts
Introduction
*Correspondence: K Imaizumi, Laboratory of Nutrition Chemistry,
Division of Bioresource and Bioenvironmental Sciences, Graduate School,
Kyushu University, Fukuoka 812-8581, Japan.
E-mail: imaizumi@agr.kyushu-u.ac.jp
Guarantor: M Iwamoto and K Imaizumi.
Contributors: MI participated in the study design, coordinated dietary
designs, carried out lipid determination, participated in the
discussion of results and wrote the paper. KI participated in the study
design, coordinated the study, participated in the discussion of results
and reviewed the paper. MS collected the data, did the statistical
analysis, participated in discussion of results and reviewed the paper.
YH participated in the study design, examined the subjects,
participated in the discussion of results and reviewed the paper. KS
participated in the study design, examined the subjects and reviewed
the paper. AT participated in the study design, coordinated the study
and participated in the discussion of results. MK carried out fatty acid
analysis and did the statistical analysis.
Received 14 May 2001; revised 30 November 2001;
accepted 3 December 2001
In a controlled, randomized, cross-over, clinical study, Sabaté
et al (1993) showed greater cholesterol-lowering effects of the
National Cholesterol Education Program Step I diet (30% fat
energy) in healthy young men, when walnuts contributed
55% fat energy than when walnuts were not supplemented
in the diet. In this study, the subjects ate all their meals in
the university kitchen. Similar studies have been done, but
the participants ate all their meals at home (Abbey et al,
1994; Chisholm et al, 1998; Zambón et al, 2000). Abbey et al
(1994) showed a cholesterol-lowering effect in an Australian
diet containing walnuts in normolipidemic men, although
only that for apolipoprotein (apo) B was statistically significant. Chisholme et al (1998) did not show the cholesterollowering effect of walnut diets in male New Zealanders. More
recently, Zamón et al (2000) confirmed cholesterol-lowering
effect in a Mediterranean diet containing walnuts in Spanish
subjects with polygenic hypercholesterolemia.
Serum lipid profiles
M Iwamoto et al
630
These studies were mainly conducted on men consuming
a Western-type diet. There have been no studies on the
effects of walnut consumption on the serum lipid profiles
in women and men eating a ‘Japanese’ diet. We thus started
to examine the effects of consuming walnuts in a carefully
controlled experimental situation as described by Sabaté et al
(1993). We measured the concentrations of serum lipids and
apolipoproteins and fatty acid composition in young women
and men on a diet based on the National Nutrition Survey of
Japan (1997) and the Recommended Dietary Allowance for
the Japanese (1995), except that walnuts contributed 12.5%
of the energy. Furthermore, we focused on the role of alinolenic acid in modifying serum cholesterol concentration,
since walnuts are a good source of this fatty acid (Dreher et al,
1996). The oxidizability of LDL was also assessed in women.
Subjects and methods
Subjects
Three investigators interviewed healthy men and women (a
total of 80) who were recruited from the Kyushu University
campus and surrounding area. They were excluded from the
trial if they ate nuts frequently, had known allergies to nuts,
smoked cigarettes at the study start, had a history of hypertension or atherosclerotic or metabolic disease, were taking
any medications regularly, or were considered unable to
comply with the study protocol. The physicians confirmed
that the participants had no cardiac, gynecologic, hepatic,
renal or endocrine disorders. After a 1.5 h orientation to the
study, the participants who met the inclusion criteria and
wanted to participate signed informed consent forms.
Twenty Japanese men and 20 women entered and completed
the experiment.
Study design
The study used a controlled, single-blind, cross-over design.
However, this study was open labeled because the subjects
and study staff could easily recognize walnuts. Every subject
followed the average Japanese diet (reference diet) during a 5
day lead-in period. After this lead-in period, all the subjects
entered into an 8 week experimental period in which they
followed each of two consecutive diets for 4 weeks. One
group consumed the walnut diet during the first period and
reference diet during the second period, while the other
group followed the diets in reverse order. A washout period
between the diets was not incorporated since a carryover
effect was not reported in the cross-over feeding study by
Sabaté et al (1993). The subjects were assigned to a particular
dietary sequence, with stratification on the basis of age,
baseline serum cholesterol concentration, and body mass
index. Ten women and 10 men consumed the walnut diet
first, and the other 20 subjects consumed the reference diet.
The subjects received all their meals at the nutritionresearch kitchen of the university during the 61 days of the
study. Breakfast and dinner had to be eaten on site in a
European Journal of Clinical Nutrition
supervised sitting. Packed lunches were provided from 11:00
to 15:00. A registered dietitian weighed and apportioned all
foods for each subject.
The subjects were told to maintain a constant level of
activity throughout the study. Compliance was assessed by
tray checks at the meals eaten on site and by self-report on
standardized forms for the packed meals. They were also
requested to record in diaries any signs of illness, medications used and any deviations from their experimental diet.
The subjects were weighed without shoes or heavy clothing at the end of the lead-in period and twice a week thereafter. Blood pressure was measured with a random-zero
sphygmomanometer at the end of the lead-in period (day
6) and twice during each experimental diet period (days 19,
33, 47 and 61). At each session, physicians measured their
blood pressures. Measurements were obtained during fasting,
before breakfast, after a 5 min rest in a sitting position.
The study protocol was approved by the University
Hospital, Faculty of Medicine, Kyushu University, in
accordance with the Helsinki Declaration of 1975 as revised
in 1983. All the subjects were offered an honorarium for
their participation.
Diets
The diets consumed during the study consisted of natural
and common Japanese food items cooked in customary ways
according to a 10 day menu cycle. The daily menus consisted
of two levels of energy intake, ranging from 10.0 to 11.1 MJ
per day for men, and from 8.37 to 9.20 MJ per day for
women, respectively. The two experimental diets were identical except that the walnut diet substituted two servings of
walnuts per day (25 or 27 g per serving, or 52 g of walnuts per
10.0 MJ) for portions of some foods in the reference diet. The
portion sizes of fatty foods, such as meat, were reduced, and
the amounts of visible fat (oils, margarine and butter) were
decreased, to accommodate the percentage of energy derived
from the walnuts (12.5%). The California Walnut Commission (Sacramento, CA, USA) donated the walnuts.
In order to design the reference diet, we chose the dietary
allowance for 20 – 39 y olds based on the Fifth Recommended
Dietary Allowance in Japan (1995) and the nutrient intake for
20 – 39 y olds based on the National Nutritional Survey of
Japan (1997). The total fat content was about 25% of the
energy value, with a proportion of saturated, monounsaturated and polyunsaturated fat of 1:1.5:1, respectively. The
reference diet had foods from all the major food groups, but
did not contain nuts, nut butters or nut oils of any kind. The
walnuts were served in several ways: mixed in salads or
cooked in dinner entrees. The walnuts gave 50, 8.5 and
12.6%, respectively, of the total fat, protein and fiber of the
walnut diet. Rice contributed 5.9, 24.8 and 15.2% of the total
fat, protein and fiber, respectively, in the walnut diet.
Complete duplicate samples of the two study diets were
obtained on 15 randomly selected days during the study
period. The samples were mixed well and analyzed for
Serum lipid profiles
M Iwamoto et al
631
macronutrients, fatty acids and sterols according to the
Official Methods of Analysis of the Official Analytical Chemists
(1974) by the Food Chemistry Center, Fukuoka, Japan. The
composition of each diet, as determined by the chemical
analyses, complied closely with the composition planned by
computer with the NUT system (version 6, Human Science
Laboratories, Shiga, Japan).
Measurements
The serum was obtained by centrifugation of blood at 4 C
(800 G, 15 min). When plasma was needed, blood was collected in vacutainers containing EDTA (4.0 mmol=l). The
serum was analyzed to determine the concentrations of
cholesterol (Allain et al, 1974) and triacylglycerols (Spayd &
Bruschi, 1974) with the use of enzyme reagent kits (Cholesterol C-test and Triglyceride G-test, respectively; Wako Pure
Chemicals Co., Tokyo, Japan). The HDL cholesterol (Warnick
et al, 1982) was measured by a commercially available kit
(HDL-C test, Wako Pure Chemicals Co., Tokyo, Japan).
The LDL cholesterol was calculated by subtraction with
the Friedewald algorithm (Friedewald et al, 1972). The concentration of the serum apo A-I and B (Ikeda et al, 1991) was
determined with a modified, commercially available turbidimetric assay (Apo A-I Auto N and Apo B Auto N; Daiichi
Pure Chemicals Co., Tokyo, Japan). The fatty acid composition of the serum cholesterol esters, phospholipids and
triacylglycerols was determined according to the methods
previously described (Ikeda et al, 1998). The determination of
the serum a-tocopherol concentration in women was done as
described previously (Zommara et al, 1998).
The plasma from the women was adjusted to a density
1.21 kg=l by addition of KBr and the LDL fraction was
obtained by density gradient ultracentrifugation (TLA100.4 rotor, Beckman, Palo Alto, CA, USA) at 543 000gmax
at 4 C for 40 min (Chung et al, 1986). This method was
chosen since it can prevent LDL from oxidation during the
isolation and the LDL are almost free of contaminating
serum protein (Sattler et al, 1994). The susceptibility of LDL
to in vitro oxidation was carried out as described by Esterbauer et al (1989). The LDL preparation was dialyzed against
a 100-fold volume of 0.01 M phosphate buffer (pH 7.4) and
0.16 M NaCl, which was made oxygen-free by vacuum degassing followed by purging with nitrogen. The LDL oxidation
was initiated by the addition of a freshly prepared aqueous
CuCl2 solution: the final conditions were 25 C, 0.25 g LDL=l
and 1.66 mM CuCl2. The oxidation resistance of LDL was
estimated in terms of the period when no oxidation occurred
(lag phase), determined as the intercept of the extrapolations
of the parts of the curve representing the lag and propagation
phases.
Statistical analysis
The values obtained at the end of the reference diet period
and walnut diet period were expressed as mean (s.e.). Accord-
ing to Fleiss (1986), a two-tailed paired t-test was done to
compare the changes in the outcome variables in response to
the dietary treatment and diet period for the two-period
cross-over design. We first examined the possible interaction
between the dietary treatment and diet period (carry-over
effect) by analyses of variance (ANOVA) and then by the twotailed paired t-test. To compare the mean values obtained
from women and men or the mean values from the different
dietary period, a two-tailed unpaired t-test was done, after
confirming the homogeneity of the variances by ANOVA.
Linear regression and Pearson correlation analyses were used
to test the correlations between the percentage change in the
serum triacylglycerols and the percentage change in the total
or LDL cholesterol concentration, and between the percentage change in a fatty acid proportion in the serum cholesteryl esters and the percentage change in LDL cholesterol
concentration.
Results
Subjects
The first week after recruitment was used to record the daily
habitual diet of the participants. Their nutrient intakes were
calculated from a computer database of foods based on the
composition of Japanese foods (NUT system; version 6,
Human Science Laboratories, Shiga, Japan; Table 1). The
subject characteristics are given in Table 2.
A man reported not eating dinner on 2 days and a woman
reported not eating breakfast or lunch for 2 days. Several
subjects reported catching a cold, feeling ill, gastric pain,
Table 1 Average daily intake of subjects for selected nutrients from 7
day diet records prior to the study
Nutrient
Women
Men
Energy (kJ=day)
Protein (percentage of energy)
Carbohydrate (percentage of energy)
Fat (percentage of energy)
Saturated (percentage of energy)
Monounsaturated (percentage of energy)
Polyunsaturated (percentage of energy)
Cholesterol (mg=day)
Dietary fiber (g=day)
Vitamin E (mg a-TE=day)
7197
13.5
52.5
31.1
11.4
13.0
6.7
334
10.8
6.9
8394
14.6
52.4
32.4
10.8
14.1
7.3
399
10.9
8.1
Table 2 Initial characteristics of the subjects. Results are expressed as
mean (s.e.) for 20 women and men, respectively
Women
Age
Body weight (kg)
Body mass index (kg=m2)
Serum cholesterol (mmol=l)
Triacylglycerol (mmol=l)
Systolic blood pressure (mmHg)
Diastolic blood pressure (mmHg)
23.6
52.0
20.7
4.53
1.51
109
66
(1.1)
(1.1)
(0.5)
(0.16)
(0.42)
(3)
(2)
Men
23.8
66.1
22.2
4.75
2.56
117
73
(0.7)
(2.0)
(0.5)
(1.90)
(0.31)
(3)
(3)
European Journal of Clinical Nutrition
Serum lipid profiles
M Iwamoto et al
632
fever and headache. Two subjects (male and female) took
medicine for a cold. None of these conditions had an impact
on the outcome of this study based on the analyses with and
without inclusion of these subjects.
Table 4 Fatty acid composition of the two diets. Values are expressed as
a percentage by weight of total fatty acid
Fatty acids
12:0
14:0
16:0
16:1
18:0
18:1
18:2
18:3
20:4
20:5
22:6
Nutrient composition of the two diets
The energy percentages from proteins, carbohydrates and
total fats, and the contents of dietary fiber, sterols and
vitamin E were maintained at a constant during both dietary
periods (Table 3). In addition, the chemical determination of
the fatty-acid content revealed the expected distribution for
each diet (Table 4). The fatty acid composition of the walnut
diet was closer to that of walnut fat, with a larger proportion
of polyunsaturated fatty acids, especially a-linolenic acid
(Dreher et al, 1996).
n-6
n-3
n-6
n-3
n-3
Reference diet
Walnut diet
0.5
1.9
19.1
2.1
8.1
42.6
19.8
3.8
0.3
0.5
1.2
0.2
0.8
12.7
1.1
5.2
25.0
44.0
10.0
0.2
0.3
0.5
Serum lipids and apoproteins
The serum total cholesterol concentrations of all the participants at the end of the lead-in period tended to be higher
than those in the first or second periods (Figure 1), presumably reflecting the differences in the dietary fat contents in
the habitual diet and the test diets (Tables 1 and 3). A crossover pattern from the first period to the second period was
observed in the values for the total and LDL cholesterol in
men and women. Since there was no evidence of a carry-over
effect between the periods, the values for serum lipids and
lipoproteins are presented in Table 5 for all the study subjects
(women and men) during each diet, irrespective of the order
of diet consumption. In women, the mean serum total
cholesterol concentration during the walnut diet period
was 0.21 mmol=l (4.9%, P ¼ 0.0038) lower than the concentration during the consumption of the reference diet. In
men, the concentration during the walnut diet was
0.16 mmol=l lower than the concentration during the reference diet (P ¼ 0.050), representing a reduction of 3.8%. The
LDL cholesterol concentration during the walnut diet period
Body weight and blood pressure
The average body weight decreased from baseline by 1.3 kg
in men and 0.1 kg in women over 54 days of the study, but
this decrease was not related to a specific diet. Subjects
during the reference diet period tended to lose weight, but
it was not statistically significant: the mean difference
between the dietary treatments in weight lost was 0.350 kg
(P ¼ 0.797) in men and 1.18 kg (P ¼ 0.102) in women, respectively. The blood pressures in men were 108=66 mmHg
during the reference diet period and 111=69 mmHg during
the walnut diet period. In women, the pressures were 101=63
and 103=63 mmHg during the reference diet period and
walnut diet period, respectively. No significant changes
were observed in the systolic (P ¼ 0.78 for women and
P ¼ 0.28 for men, respectively) or diastolic (P ¼ 0.91 for
women and P ¼ 0.291 for men, respectively) blood pressures
at the end of each diet period.
Table 3 Nutrient composition of the two study diets; values analyzed during chemical analysis of samples from the
study diets
Reference diet
Nutrient
Energy (kJ=day)
Protein (percentage of energy)
Carbohydrate (percentage of energy)
Fat (percentage of energy)
Saturated (percentage of energy)
Monounsaturated (percentage of energy)
Polyunsaturated (percentage of energy)
Cholesterol (mg=day)
Campesterol (mg=day)
Stigmasterol (mg=day)
b-sitosterol (mg=day)
Dietary fiber (g=day)
Vitamin E (mg a-TE=day)
European Journal of Clinical Nutrition
Walnut diet
Calculated
Analyzed
Calculated
Analyzed
10081
14.6
60.0
25.3
7.2
10.8
7.4
395
9054
13.8
62.4
24.0
6.9
10.3
6.8
279
46.2
9.6
89.4
10122
14.2
60.0
25.7
5.2
6.9
13.9
314
9684
13.6
59.6
26.0
4.8
6.7
14.5
252
23.0
7.5
90.4
19.5
10.4
22.3
9.1
Serum lipid profiles
M Iwamoto et al
633
Figure 1 The concentrations of serum total, LDL and HDL cholesterol in women and men in the lead-in, first and second periods. Open bars show the
subjects who consumed the walnut diet first and the reference diet in the second period, hatched bars show the subjects who consumed the diet in
reverse order. The upper and lower panels show the lipid concentrations for women and men, respectively. Asterisks indicate significant differences
between walnut and reference diets at *P ¼ 0.10 and **P ¼ 0.08, respectively. Values are expressed as mean (s.e.) for 10 subjects.
Table 5 Serum lipid and lipoprotein concentrations of women, men and all subjects and the difference between
reference diet and walnut diet period. Reference diet values were subtracted from walnut diet values according to
methods described by Fleiss (1986). Results are expressed as mean (s.e.) for 20 women and 20 men and all subjects,
respectively
Variable
Reference diet
Walnut diet
Change
Total cholesterol (mmol=l)
Women
Men
All
4.26 0.17
4.17 0.13
4.22 0.10
4.05 0.17
4.00 0.16
4.03 0.11
7 0.21 0.07
7 0.16 0.08
7 0.19 0.05
7 4.9
7 3.8
7 4.5
0.007
0.054
0.001
LDL cholesterol (mmol=l)
Women
Men
All
2.08 0.13
2.02 0.14
2.05 0.09
1.87 0.13
1.85 0.14
1.86 0.09
7 0.22 0.06
7 0.18 0.09
7 0.20 0.05
7 10.6
7 8.9
7 9.8
0.001
0.078
0.001
HDL cholesterol (mmol=l)
Women
Men
All
1.68 0.06
1.32 0.06
1.50 0.05
1.65 0.06
1.32 0.06
1.48 0.05
7 0.03 0.04
0.00 0.02
7 0.02 0.02
7 1.8
0
7 1.3
0.395
0.909
0.490
LDL cholesterol:HDL cholesterol
Women
Men
All
1.28 0.08
1.65 0.17
1.46 0.10
1.13 0.07
1.49 0.17
1.31 0.09
7 0.11 0.05
7 0.16 0.06
7 0.15 0.04
7 8.6
7 9.7
7 10.3
0.009
0.025
0.001
Triacylglycerol (mmol=l)
Women
Men
All
1.13 0.08
1.86 0.14
1.50 0.10
1.16 0.07
1.83 0.15
1.50 0.10
0.03 0.09
7 0.03 0.13
0.00 0.08
2.7
7 1.6
0
0.729
0.824
0.993
Apolipoprotein A I (mg=l)
Women
Men
All
1474 50
1295 38
1385 34
1425 36
1279 40
1352 29
7 50 39
7 16 20
7 33 22
7 3.4
7 1.2
7 2.4
0.220
0.434
0.142
698 37
759 44
728 29
636 35
708 46
672 29
7 62 13
7 51 21
7 56 12
7 8.9
7 6.7
7 7.7
0.001
0.026
0.001
Apolipoprotein B (mg=l)
Women
Men
All
%D
P-value
European Journal of Clinical Nutrition
Serum lipid profiles
M Iwamoto et al
634
in women was 0.22 mmol=l lower than the concentration
during the reference diet period (P ¼ 0.0008), representing a
reduction of 11%; however, this LDL cholesterol reduction
was not significant in men (P ¼ 0.078). There were no significant differences in the serum HDL cholesterol or triacylglycerols in men or women due to diet. The men had a
higher serum triacylglycerol concentration than the women,
while it was the reverse for the HDL cholesterol. The ratios
of LDL to HDL cholesterol in men and women were
significantly lower during the walnut diet period than
during the reference diet period (P ¼ 0.0088 for women
and P ¼ 0.025 for men, respectively). The serum apo B
concentrations in women and men were lower during
consumption of the walnut diet than of the reference
diet (P ¼ 0.00017 and P ¼ 0.026 for men and women,
respectively). The concentration of the serum apo A-I was
not influenced by diet.
Serum lipids fatty acids
The fatty acid compositions of cholesteryl esters in women
and men during the reference diet were as follows (mol%,
mean (s.e.)): 14:0, 1.8 (0.2) and 0.9 (0.2); 16:0, 10.4 (0.3) and
10.8 (0.2); 16:1, 2.0 (0.1) and 2.0 (0.1); 18:0, 0.9 (0.1) and 1.2
(0.1); 18:1, 19.3 (0.3) and 21.0 (0.3); 18:2, 57.0 (0.8) and 54.2
(0.7); 18:3, 0.5 (0.0) and 0.4 (0.0); 20:4, 6.3 (0.3) and 7.6
(0.5); 20:5, 1.3 (0.2) and 1.2 (0.1); 22:6, 0.6 (0.1) and 0.8 (0.1)
for women and men, respectively. Both women and men on
the walnut diet had a significantly increased proportion of alinolenic acid and linoleic acid in the serum at the expense
of palmitic, palmitoleic, oleic and arachidonic acids (Figure
2). In particular, a-linolenic acid showed the most prominent
change after the walnut diet (76.3%, P < 0.0001 for women
Figure 2 The percentage change in fatty acid composition of the serum
cholesteryl esters in women (filled bar) and men (open bar) from the
reference diets. Asterisks indicate significant difference between walnut
and reference diets at *P < 0.05, **P < 0.01. Values are expressed as
mean (s.e.) for 20 subjects.
European Journal of Clinical Nutrition
and 107%, P < 0.0001 for men, respectively). There were no
significant diet effects on the proportion of eicosapentaenoic
and docosahexaenoic acids (P > 0.13). Effects similar to those
for cholesteryl esters were seen for the fatty acid composition
of the serum triacylglycerols and phospholipids, except there
were no significant differences in the proportion of a-linolenic acid in the phospholipids in women (P ¼ 0.43), and
arachidonic acid in the phospholipids and triacylglycerols in
men (P > 0.29). These fatty acid changes confirm that
subjects adhered closely to the diet (Table 4).
The relationship between serum cholesterol and
triacylglycerols or fatty acid of cholesteryl esters in
individual subjects
The relationship between the serum LDL cholesterol concentration and the triacylglycerol concentration in individual subjects is shown in Figure 3 as their percentage change
when the walnut and the reference diets were replaced. Half
of the women increased or decreased their serum triacylglycerol concentrations after the walnut diet. As a whole, there
was a negative correlation between the changes in the
triacylglycerol and LDL cholesterol concentrations, but it
was not statistically significant (r ¼ 0.266, P ¼ 0.227). Ten
out of 20 men had increased serum triacylglycerol concentrations after the walnut diet. As a whole, there was a
significantly negative correlation between the changes in
the triacylglycerol and LDL cholesterol concentrations
(r ¼ 0.636, P ¼ 0.0018). When this calculation was done
with women and men together, there was a significant
dependency of the change in the triacylglycerols on the
change in the LDL cholesterol concentrations (r ¼ 0.466,
P ¼ 0.0069).
The relationship between the LDL cholesterol and a-linolenic acid of serum cholesteryl esters was measured as the
percentage change when the walnut and reference diets were
replaced. All the subjects had increased proportions of alinolenic acid after the walnut diet. In women, there was a
significantly negative correlation between the changes in the
Figure 3 The relationship between serum triacylglycerols and LDL
cholesterol concentration in woman (filled circle) and man (cross) as
percentage change from the reference diets.
Serum lipid profiles
M Iwamoto et al
635
a-linolenic acid and the LDL cholesterol (r ¼ 0.496,
P ¼ 0.019). In men, there was a negative correlation between
the a-linolenic acid and LDL cholesterol change, but it did
not reach a significant level (r ¼ 0.326, P ¼ 0.138). Besides alinolenic acid, there was no clear correlation between the
percentage changes in LDL cholesterol concentration and
the percentage changes in linoleic acid (r < 0.05), eicosapentaenoic acid (r < 0.18) or docosahexaenoic acid (r < 0.15) of
cholesteryl ester in men and women.
No significant correlation was found between the percentage changes in the serum total cholesterol concentrations
and the percentage change in a-linolenic acid (r ¼ 0.109,
P ¼ 0.636 for women and r ¼ 0.261, P ¼ 0.236 for men,
respectively) or linoleic acid (r ¼ 0.170, P ¼ 0.443 for
women and r ¼ 0.261, P ¼ 0.236 for men, respectively).
LDL oxidizability
In women, there were no significant differences in the lag
times (resistance of LDL to Cu2þ-induced oxidation) between
the reference and walnut diet groups (39.4 (0.9) and 37.0
(0.5) min for reference diet and walnut diet, respectively,
P ¼ 0.19). The serum a-tocopherol concentration in women
was similar during the reference diet and walnut diet period
(21.2 (2.3) and 20.0 (1.7) mmol=l during reference and walnut
diet period, respectively, P ¼ 0.63).
Discussion
These results from a controlled, randomized, cross-over
experiment showed that replacing 55% of the fat energy in
a reference diet (Japanese diet with 26% fat energy) with
walnuts lowered the serum cholesterol concentrations by 5%
for normal women (P < 0.01) and 4% for normal men
(P ¼ 0.05), respectively. The effects of the walnut diet on
the LDL cholesterol concentration were also greater in
women than in men. It should be noted that the walnut
diet resulted in lowering the serum apo B and the ratio of
LDL cholesterol to HDL cholesterol in women and men in
comparison with the reference diet (P < 0.05). In addition,
the HDL cholesterol concentrations were not changed in
both sexes. Furthermore, the walnut diet had no significant
effect on the blood pressure (P > 0.4). Our results essentially
are consistent with the well-controlled diet study reported by
Sabaté et al (1993), who detected 12.4% reduction in the
total cholesterol in healthy American men, when walnuts
replaced 55% of the fat energy in a National Cholesterol
Education Program Step I diet (30% fat energy) for 4 weeks.
Unlike our study, they found a significant reduction in the
LDL cholesterol (16.3%) and HDL cholesterol (4.9%) in men
after supplementation with walnuts. The hypocholesterolemic effects of walnut consumption have also been reported
in randomized, cross-over trials with free-living healthy
Australian men (Abbey et al, 1994) and free-living hypercholesterolemic Spanish men and women (Zamón et al, 2000).
The latter study did not analyze the results of women and
men separately. In the Australian trial (35.7% fat energy),
50% of the fat energy was replaced with walnuts for 3 weeks.
In the Spanish trial (33% fat energy), 35% of the fat energy
was replaced with walnuts for 6 weeks. Taken together, these
results indicate that walnut supplementation at 35 – 55% of
fat energy to the currently recommended cholesterol lowering diets (Japanese, Mediterranean and National Cholesterol Education Program Step I diet) which supply 26 – 33%
fat energy exert beneficial effects on lipoprotein risk profiles
despite racial and sex differences.
In well-controlled studies with a large number of subjects
(Ginsberg et al, 1998; Howard et al, 1995; Mensink & Katan,
1990), they showed that diet effects were similar in women
and men when they investigated the effects of reducing the
dietary saturated fatty acids on serum total and LDL cholesterol concentrations. In this study, the ratio of polyunsaturated to saturated fat was 3.0 in the walnut diet and 1.0 in
the reference diet. The estimated effect of the walnut diet on
the LDL cholesterol concentration tended to be similar in
both women and men, but it was not identical. It could be
expected that the observed differences in this study would
have been diminished if the number of subjects had been
increased. However, the analyses based on Figure 3 allowed
us to explain in part the different responses between both
sexes. The LDL cholesterol reduction after the walnut diet
was independent of an increase or a decrease in the triacylglycerol concentration in women (P > 0.2), whereas it was
dependent on the triacylglycerol elevation (P < 0.01). Therefore, these results might suggest that, in men, the LDL
cholesterol-lowering effect of a walnut-containing diet intimately involves their own triacylglycerol metabolism. This
may not be the case for women.
Fatty acids in serum cholesteryl esters are considered to be
quantitative biomarkers of long-term dietary intake in
humans (Zock et al, 1997). Both a-linolenic acid and linoleic
acid increased after the walnut diet, but the extent was
greater for a-linolenic acid (76.3 and 107% for women and
men, respectively) than for linoleic acid (7.4 and 9.4% for
women and men, respectively). The percentage change in
the a-linolenic acid after the walnut diet negatively correlated to the percentage change in the LDL cholesterol concentration in the subjects, particularly in women. This
relationship was not found for linoleic acid and the desaturation products of a-linolenic acid (eicosapentaenoic and
docosahexaenoic acid). These results might imply that the
lowering of LDL cholesterol by the walnut diet was attributed to the a-linolenic acid. Alternatively, from the analyses
of Figure 3, a-linolenic might be involved in preventing the
conversion of VLDL to LDL. In any event, these results agree
with the previous findings that the consumption of a-linolenic acid has a lowering effect on the serum cholesterol
concentration where there are considerable amounts of alinolenic acid available (Chan et al, 1991; Valsta et al, 1995).
It remains a possibility that dietary components such as
cholesterol (Zanni et al, 1987), plant sterols (Hallikainen et
al, 2000) and dietary fibers (Jenkins et al, 2000) contributed
European Journal of Clinical Nutrition
Serum lipid profiles
M Iwamoto et al
636
to the observed differences between the reference diet and
the walnut diet. However, as shown in Table 3, both diets
contained comparable amounts of these components. Walnuts are a rich source of arginine (Dreher et al, 1996) and the
low ratio of lysine to arginine may also have some effect on
the serum cholesterol (Kritchevsky et al, 1982). This would
not be the case here, as both diets contained comparable
amounts of rice, which is composed of hypocholesterolemic
proteins with a low ratio of lysine to arginine (Morita et al,
1997). Based on a computer database of foods (NUT system;
version 6, Human Science Laboratories, Siga, Japan), the rice
supplemented to the diets should have contributed approximately 25% of the total protein of both diets, whereas the
walnut protein was 8.5% of the walnut diet.
Three prospective cohort studies (Fraser et al, 1992; Prineas et al, 1993; Hu et al, 1998) have documented that
frequent nut consumption is associated with a reduced risk
of both total coronary heart disease and non-fatal myocardial infarction. Although classifying consumption patterns
of specific nuts in these studies is difficult, these beneficial
effects could be due in part to the effects of walnut consumption on the lipoprotein profiles reflected in the lowering of total and LDL cholesterol and the ratio of LDL=HDL
cholesterol (Sabaté & Fraser, 1994). In addition to improving
the serum cholesterol profiles, it is likely that the serum fatty
acid profiles of the walnut diets could contribute to the
reduced risk of both coronary heart disease and non-fatal
myocardial infarction. This hypothesis is supported by
results of the multiple risk factors intervention trial (Dolecek, 1992), and a secondary prevention trial in France (de
Lorgeril et al, 1994, 1999) and a cohort study of women in
the USA (Ascherio et al, 1996). All these studies have shown
that there is an inverse association between the intake of alinolenic acid and the risk of coronary disease. In addition to
an elevated proportion of a-linolenic acid, we showed a
decreased proportion of arachidonic acid ( 7 17.5% for
women and 7 22.4% for men, respectively) after the
walnut diet. This effect could be ascribed to the role of alinolenic acid, which inhibits D6-desaturase activity in
experimental animals (Garg et al, 1988).
It has been reported that diets high in linoleic acid result
in LDL that are more susceptible to in vitro oxidation than
those high in oleic acid (Bonanome et al, 1992; Reaven et al,
1991). Therefore, we measured the susceptibility of LDL to
copper-mediated oxidation in women fed on reference and
walnut diets. The results showed that there was no walnutdependent effect on the lag time, when the walnut diet and
reference diet supplied similar levels of vitamin E (Table 3).
Furthermore, the serum a-tocopherol concentration in
women was similar during the reference and walnut diet
period. These results are in agreement with the report by
Zamón et al, (2000), who found that the lag time during the
copper-induced LDL oxidation in their subjects was similar
during the reference diet and walnut diet.
In conclusion, these results indicate that a diet that
includes moderate quantities of walnuts, without an overall
European Journal of Clinical Nutrition
increase in total dietary fat and energy, lowers serum cholesterol concentration and favorably modifies the serum
lipoprotein and fatty acid profile in normal Japanese
women and men, without influencing the LDL oxidizability.
They also suggest that a-linolenic acid in the walnut diet
appears to be responsible for lowering LDL cholesterol,
particularly in women.
References
Abbey M, Noakes M, Belling GB & Nestel PJ (1994): Partial replacement of saturated fatty acids with almonds or walnuts lowers total
plasma cholesterol and low-density-lipoprotein cholesterol. Am. J.
Clin. Nutr. 59, 995 – 999.
Allain CC, Poon LS, Chan CSG, Richmond W & Fu PC (1974):
Enzymatic determination of total serum cholesterol. Clin. Chem.
20, 470 – 475.
Ascherio A, Rimm EB, Giovannucci EL, Spiegelman D, Stampfer M &
Willet WC (1996): Dietary fat and risk of coronary heart disease in
men: cohort follow up study in the United States. Br. Med. J. 313,
84 – 90.
Bonanome A, Pagnan A, Biffanti S, Opportuno A, Sorgato F, Dorella
M, Maiorino M & Ursini F (1992): Effect of dietary monounsaturated and polyunsaturated fatty acids on the susceptibility of
plasma low density lipoproteins to oxidative modification. Arteriosclerosis 12, 529 – 533.
Chan JK, Bruce VM & McDonald BE (1991): Dietary a-linolenic acid is
as effective as oleic acid and linoleic acid in lowering blood
cholesterol in normolipidemic men. Am. J. Clin. Nutr. 53, 1230 –
1234.
Chisholm A, Mann J, Skeaff M, Frampton C, Sutherland W, Duncan
A & Tiszavari S (1998): A diet rich in walnuts favorably influences
plasma fatty acid profile in moderately hyperlipidaemic subjects.
Eur. J. Clin. Nutr. 52, 12 – 16.
Chung BH, Segrest JP, Ray MJ, Brunzell JD, Hokanson JE, Krauss RM,
Beaudrie K & Cone JT (1986): Single vertical density gradient
ultracentrifugation. Meth Enzymol. 128, 181 – 209.
de Lorgeril M, Renaud S, Mamelle N, Salen P, Martin JL, Monjaud I,
Guidollet J, Touboul P & Delaye J (1994): Mediterranean alphalinolenic acid-rich diet in secondary prevention of coronary heart
disease. Lancet 343, 1454 – 1459.
de Lorgeril M, Salen P, Martin J-L, Monjaud I, Delaye J & Mamelle N
(1999): Mediterranean diet, traditional risk factors, and the rate of
cardiovascular complications after myocardial infarction: final
report of the Lyon Diet Heart Study. Circulation 99, 779 – 785.
Dolecek TA (1992): Epidemiological evidence of relationships
between dietary polyunsaturated fatty acids and mortality in the
multiple risk factor intervention trial. Proc. Soc. Exp. Biol. Med. 200,
177 – 182.
Dreher ML, Maher CV & Kearney P (1996): The traditional and
emerging role of nuts in healthful diets. Nutr. Rev. 54, 241 – 245.
Esterbauer H, Striegl G, Puhl H & Rotheneder M (1989): Continuous
monitoring of in vitro oxidation of human low density lipoprotein. Free Rad. Res. Commun. 6, 67 – 75.
Fleiss JL (1986): The Design and Analysis of Clinical Experiments. New
York: John Wiley.
Fraser GE, Sabaté J, Beeson WL & Strahan TM (1992): A possible
protective effect of nut consumption on risk of coronary heart
disease. Arch. Intern. Med. 152, 1416 – 1424.
Friedewald WT, Levy RI & Fredrickson DS (1972): Estimation of the
concentration of low-density lipoprotein cholesterol in plasma,
without use of the preparative ultracentrifuge. Clin. Chem. 18,
499 – 502.
Garg J, Sebokova E, Thomson BR & Clandinin T (1988): D6-Desaturase activity in liver microsomes of rats fed diets enriched with
cholesterol and=or o3 fatty acids. Biochem. J. 249, 351 – 356.
Serum lipid profiles
M Iwamoto et al
Ginsberg HN, Kris-Etherton P, Dennis B, Elmer PJ, Ershow A, Lefevre
M, Pearson T, Roheim P, Ramakrishnan R, Reed R, Stewart K,
Stewart P, Phillips K & Anderson N (1998): Effects of reducing
dietary saturated fatty acids on plasma lipids and lipoproteins in
healthy subjects. the Delta Study, Protocol 1. Arterioscler. Thromb.
Vasc. Biol. 18, 441 – 449.
Hallikainen MA, Sarkkinen ES & Uusitupa MIJ (2000): Plant stanol
esters affect serum cholesterol concentrations of hypercholesterolemic men and women in a dose-dependent manner. J. Nutr. 130,
767 – 776.
Howard BV, Hannah JS, Heiser CC & Jablonski KA (1995): Effects of
sex and ethnicity on responses to a low-fat diet: a study of African
Americans and whites. Am. J. Clin. Nutr. 62, 488S – 492S.
Hu FB, Sampfer MJ, Manson JE, Rimm EB, Colditz GA, Rosner BA,
Speizer FE, Hennekens CH & Willett WC (1998): Frequent nut
consumption and risk of coronary heart disease in women: prospective cohort study. Br. Med. J. 317, 1341 – 1345.
Ikeda T, Shibuya Y, Senba U, Sugiuchi H, Araki S, Uji Y & Okabe H
(1991): Automated immunoturbidimetric analysis of six plasma
apolipoproteins: correlation with radial immunodiffusion assays.
J. Clin. Lab. Anal. 5, 90 – 95.
Ikeda I, Yoshida H, Tomooka M, Yosef A, Imaizumi K, Tsuji H & Seto
A (1998): Effects of long-term feeding of marine oils with different
positional distribution of eicosapentaenoic and docosahexaenoic
acids on lipid metabolism, eicosanoid production, and platelet
aggregation in hypercholesterolemic rats. Lipids 33, 897 – 904.
Jenkins DJA, Cyril Kendall CWC, Axelsen M, Augustin SA & Vuksan
V (2000): Viscous and nonviscous fibres, nonabsorbable and low
glycaemic index carbohydrates, blood lipids and coronary heart
disease. Curr. Opin. Lipidol. 11, 49 – 56.
Kritchevsky D, Tepper SA, Czarnecki SK & Klurfeld DM (1982):
Atherogenicity of animal and vegetable protein: influence of the
lysine to arginine ratio. Atherosclerosis 41, 429 – 431.
Mensink RP & Katan MB (1990): Effect of a diet enriched with
monounsaturated or polyunsaturated fatty acids on levels of
low-density and high-density lipoprotein cholesterol in healthy
women and men. New Engl. J. Med. 322, 402 – 404.
Morita T, Oh-hashi A, Takei K, Ikai M, Kasaoka S & Kiriyama S (1997):
Cholesterol-lowering effects of soybean, potato and rice proteins
depend on their low methionine contents in rats fed a cholesterolfree purified diet. J. Nutr. 127, 470 – 477.
National Nutrition Survey of Japan (1997): Annual Report: Health Service
Bureau, Ministry of Health and Welfare. Tokyo: Daiichi Shuppan.
Official Methods of Analysis of the Association of Official Analytical
Chemists (1987): 15th edn. Arlington, VA: Association of Official
Analytical Chemists.
Prineas RJ, Kushi LH, Folsom AR, Bostick RM & Wu Y (1993): Walnuts
and serum lipids. New Engl. J. Med. 328, 603 – 607.
Reaven P, Parthasarathy S, Grasse BJ, Miller E, Almazan F, Mattson
FH, Khoo JC, Steinberg D & Witztum JL (1991): Feasibility of using
an oleate-rich diet to reduce the susceptibility of low-density
lipoprotein to oxidative modification in humans. Am. J. Clin.
Nutr. 54, 701 – 706.
Recommended Dietary Allowance for the Japanese (1995): 5th revision.
Health Service Bureau, Minstry of Health and Welfare, Japan.
Tokyo: Daiichi Shuppan.
Sabaté J & Fraser GE (1994): Nuts: a new protective food against
coronary heart disease. Cur. Opin. Lipidol. 5, 11 – 16.
Sabaté J, Fraser GE, Burke K, Knutsen SF, Bennett H & Lindsted KD
(1993). Effects of walnuts on serum lipid levels and blood pressure
in normal men. New Engl. J. Med. 328, 603 – 607.
Sattler W, Mohr D & Stocker R (1994): Rapid isolation of lipoproteins
and assessment of their peroxidation by high-performance liquid
chromatography postcolumn chemiluminescence. Meth. Enzymol.
233, 469 – 489.
Spayd RW & Bruschi B (1974): Multilayer film elements for clinical
analysis applications to representative chemical determinations.
Clin. Chem. 124, 1343 – 1350.
Valsta LM, Jauhiainen M, Aro A, Salminen I & Mutanen M (1995):
The effects on serum lipoprotein levels of two monounsaturated
fat rich diets differing in their linoleic and a-linolenic acid contents. Nutr. Metabl. Cardiovasc. Dis. 5, 129 – 140.
Warnick GR, Benderson J & Albers JJ (1982): Dextran sulfate-Mg2þ
precipitation procedure for quantitation of high density lipoprotein cholesterol. Clin. Chem. 28, 1379 – 1388.
Zambón D, Sabaté J, Muñoz S, Campero B, Casals E, Merlos M,
Laguna JC & Ros E (2000): Substituting walnuts for monounsaturated fat improves the serum lipid profile of hypercholesterolemic
men and women. a randomized crossover trial. Ann. Intern. Med.
132, 538 – 546.
Zanni EE, Zannis VI, Blum CB, Herbert PN & Breslow JL (1987): Effect
of egg cholesterol and dietary fats on plasma lipids, lipoproteins,
and apoproteins of normal women consuming natural diets. J.
Lipid Res. 28, 518 – 527.
Zock PL, Mensink RP, Harryvan J, de Vries JH & Katan MB (1997):
Fatty acids in serum cholesteryl esters as quantitative biomarkers
of dietary intake in humans. Am. J. Epidemiol. 145, 1114 – 1122.
Zommara M, Toubo H, Sakono M & Imaizumi K (1998): Prevention
of peroxidative stress in rats fed on a low vitamin E-containing
diet by supplementing with a fermented bovine milk whey preparation: effect of lactic acid and b-lactoglobulin on antiperoxidative action. Biosci. Biotechnol. Biochem. 62, 710 – 717.
637
European Journal of Clinical Nutrition