Effects of a high walnut and high cashew nut diet on selected markers
of the metabolic syndrome: a controlled feeding trial
Janine Mukuddem-Petersen1,2*, Welma Stonehouse (Oosthuizen)1,3, Johann C. Jerling1,
Susanna M. Hanekom1and Zelda White1
1School of Physiology, Nutrition and Consumer Science, North-West University (Potchefstroom Campus), Potchefstroom,
2School of Computer, Mathematical and Statistical Sciences, North-West University (Potchefstroom Campus), Potchefstroom,
3Institute of Food, Nutrition and Human Health, Massey University (Albany Campus), North Shore, Auckland, New Zealand
(Received 18 July 2006 – Revised 20 December 2006 – Accepted 3 January 2007)
We investigated the effects of a high walnut diet and a high unsalted cashew nut diet on selected markers of the metabolic syndrome. In a ran-
domized, parallel, controlled study design, sixty-four subjects having the metabolic syndrome (twenty-nine men, thirty-five women) with a mean
age of 45 (SD 10) years and who met the selection criteria were all fed a 3-week run-in control diet. Hereafter, participants were grouped according
to gender and age and then randomized into three groups receiving a controlled feeding diet including walnuts, or unsalted cashew nuts or no nuts
for 8 weeks. Subjects were required to have lunch at the metabolic ward of the Nutrition Department of the North-West University (Potchefstroom
Campus). Both the walnut and the unsalted cashew nut intervention diets had no significant effect on the HDL-cholesterol, TAG, total cholesterol,
LDL-cholesterol, serum fructosamine, serum high-sensitivity C-reactive protein, blood pressure and serum uric acid concentrations when com-
pared to the control diet. Low baseline LDL-cholesterol concentrations in the cashew nut group may have masked a possible nut-related benefit.
Plasma glucose concentrations increased significantly (P¼0·04) in the cashew nut group compared to the control group. By contrast, serum fruc-
tosamine was unchanged in the cashew nut group while the control group had significantly increased (P¼0·04) concentrations of this short-term
marker of glycaemic control. Subjects displayed no improvement in the markers of the metabolic syndrome after following a walnut diet or a
cashew nut diet compared to a control diet while maintaining body weight.
Metabolic syndrome: Hyperlipidaemia: Hypertension: Obesity: Lipids: Fatty acids: Diet
The National Cholesterol Education Program’s Adult Treat-
ment Panel III (ATP III) defines the metabolic syndrome by
easily measured clinical parameters that include: increased
abdominal circumference, elevated TAG, low HDL-choles-
terol (HDL-C), elevated fasting blood glucose and/or elevated
blood pressure (BP). Three of these five are required for diag-
nosis (National Cholesterol Education Program, 2001).
From the third National Health and Nutrition Examination
Survey (NHANES III) (Ford, 2004) and the THUSA study
(THUSA is the acrynom for Transition and Health during
Urbanisation in South Africa, and is also the Tswana word
for “help”) (Oosthuizen et al. 2002) it is evident that the
increasing prevalence of the metabolic syndrome is a health
problem not only for developed countries but also for develop-
This syndrome has been identified as a target for dietary
therapies (Wirfalt et al. 2001) to reduce the risk of CVD
(Shirai, 2004) and type 2 diabetes (Jimenez-Cruz et al.
2003). In lieu of the therapeutics of the metabolic syndrome
approaches. Some of the recent dietary interventions range
from caloric restriction (Wien et al. 2003; Muzio et al.
2005), type of diet, e.g. Mediterranean (Esposito et al. 2004;
Michalsen et al. 2006), type and/or amount of dietary fat (Ric-
cardi et al. 2004; Freire et al. 2005), amount of carbohydrates
(Dansinger & Schaefer, 2006), increasing the intake of certain
minerals (He et al. 2006) and increasing dietary intake of cer-
tain functional foods (e.g. nuts, grapefruit) (Pieters et al. 2005;
Fujioka et al. 2006).
Riccardi et al. (2004) concluded that the diet for the treat-
ment of the metabolic syndrome should be limited in the
intake of saturated fat, while high-fibre/low-glycaemic index
foods should be used without specific limitations (Giacco
et al. 2000). Recently, in the Framingham Offspring cohort,
the researchers found glycaemic index and glycaemic load
(a measure of both carbohydrate quality and quantity) were
*Corresponding author: Dr Janine Mukuddem-Petersen, School of Computer, Mathematical and Statistical Sciences, North-West University (Potchefstroom
Campus), Private Bag X 6001, Potchefstroom, South Africa, fax þ27 18 299 2570, email Janine.MukuddemPetersen@NWU.ac.za
Abbreviations: ATP III, National Cholesterol Education Program’s Adult Treatment Panel III; BP, blood pressure; %E, percentage of total energy; HDL-C,
HDL-cholesterol; LDL-C, LDL-cholesterol; S-hs CRP, serum high-sensitivity C-reactive protein; TC, total cholesterol.
positively associated with the prevalence of the metabolic syn-
drome (McKeown et al. 2004).
Epidemiological findings indicate that frequent nut con-
sumption offers protection from fatal and non-fatal CHD
events (Sabate, 1993). Nuts have a low glycaemic index
(Foster-Powell et al. 2002) and are a rich source of protein,
unsaturated fatty acids (MUFA and PUFA), vitamin E, B6,
folic acid and niacin, fibre, magnesium, potassium, arginine,
phytosterols and other phytochemical compounds (such as
polyphenols and ellagic acid) (Dreher et al. 1996). The protec-
tive effects of nuts are mediated through several mechanisms.
Specifically, the health benefits of walnuts include lowering
cholesterol, increasing the ratio of HDL-C to total cholesterol
(TC), decreasing TAG and increasing HDL-C, reducing
inflammation and improving arterial function in patients
with type 2 diabetes and hyperlipidaemia (Ros et al. 2004;
Tapsell et al. 2004; Zhao et al. 2004; Zibaeenezhad et al.
2005). Besides our recent contribution (Pieters et al. 2005),
there is a scarcity of literature on the effects of cashew nut
In fact, to date there are no other studies that investigated
the effects of nuts on either markers of the metabolic syn-
drome or using subjects with metabolic syndrome. Therefore,
the primary objective of the present study was to determine
the effects of a high walnut diet and a high unsalted cashew
nut diet on markers of this syndrome compared to a control
diet. In this contribution we report the effects on the following
markers: serum lipids (TC, LDL-cholesterol (LDL-C), HDL-C
serum uric acid and serum high-sensitivity C-reactive protein
Subjects and methods
All participants gave written informed consent and the study
was approved by the Ethics Committee of the Potchefstroom
Campus of the North-West University. Sixty-eight White/Cau-
casian volunteers with the metabolic syndrome were recruited
mainly from the Potchefstroom Campus of the North-West
University and surrounding areas in Potchefstroom, South
Africa. According to the power calculations providing 80%
power at 5% significance based on LDL-C data from several
studies (Chisholm et al. 1998; Jenkins et al. 2002; Lovejoy
et al. 2002; Sabate et al. 2003; Ros et al. 2004), twenty sub-
jects per treatment group were needed to detect a decrease of
15% (0·6mmol/l) which might reduce the risk of CVD by
more than 20% (Moruisi et al. 2006). The ATP III criteria
for the diagnosis of the metabolic syndrome was used. In
this regard, the metabolic syndrome was defined as the pre-
sence of three or more of the following criteria (Ford et al.
2002): abdominal obesity (waist circumference .88cm for
women or .102cm for men); fasting TAG $ 1·7mmol/l;
HDL-C # 1·0mmol/l for men, HDL-C # 1·3mmol/l for
women; BP $ 130/85mmHg (the use of anti-hypertensive
medication was also an indication of high BP) and fasting glu-
cose $6·1mmol/l (fasting finger prick blood glucose concen-
trations were measured with a SureStepe blood glucose meter
(Lifescan Inc., Milpitas, CA, USA), using Fine Point lancets,
and SureStepe test strips (code 11).
Additional inclusion criteria included subjects being able to
comply with controlled feeding conditions; being willing and
able to eat walnuts and cashew nuts, and participants had to be
older than 21 and younger than 65 years. Pregnancy or lacta-
tion, thiazide (.25mg/d) and b-blocker (non-specific, b1and
b2)use, subjects having nut allergies and diagnosed diabetes
formed part of the exclusion criteria.
A randomized, controlled, parallel, study design was used.
The study protocol consisted of a 3-week run-in period
during which the subjects consumed a control diet (percentage
of total energy (%E) intake from protein–carbohydrate–
fat ¼ 20:47:33; Table 1). After the run-in period participants
were grouped according to gender and age and then into
three groups by randomly drawing numbers from a hat.
Group one received walnuts (n 21), group two received
unsalted cashew nuts (n 21), while group three continued
with the control diet without any nuts or nut-based ingredients
(n 22). Furthermore, these three intervention diets were fol-
lowed for 8 weeks. Due to practical reasons the study was
divided into three cohorts distributed over a 1-year period.
All the food was provided to the participants for the duration
Table 1. Planned and analysed composition of diets as well as the habitual diets
Nutrients, glycaemic index and
Habitual diet prior to dietary
intervention (n 64)
Walnut diet (n 21)Cashew nut diet (n 21)Control diet (n 22)
Protein (% of energy)
Carbohydrate (% of energy)
Fat (% of energy)
SFA (% of energy)
MUFA (% of energy)
PUFA (% of energy)
†Determined by using the FoodFinder 2 Program (Medical Research Council of South Africa, Tygerberg; mean of 14d menu).
§Calculated with the aid of the glycaemic index and glycaemic load tables of Foster-Powell et al. (2002).
of the trial. Fasting blood samples, oral glucose tolerance tests,
anthropometric measurements and BP measurements were
taken before (after the 3-week run-in period) and after
the intervention period (8-week controlled feeding). BMI
(kg/m2) was calculated. In addition, the weighing of subjects
was done twice weekly throughout the run-in period and the
experimental phase. The subjects were informed of all aspects
of the study before commencement.
The proportion of total energy (ranging from 63 to 108g/d)
from nuts was 20%. Except for the nuts, the diets were iden-
tical. This was achieved by making proportional reductions to
all food portions in the walnut and the unsalted cashew nut
diet menus to accommodate the energy supplied by the
respective nuts (Tables 1 and 2). The study featured a
highly controlled feeding protocol. In this regard, all subjects
were required to have their lunch at the metabolic ward of the
Department of Nutrition at the Potchefstroom Campus of the
North-West University. Breakfast and dinner were provided
in take-away format. Using pre-packed food parcels and a var-
iety of set menu options ensured compliance. On a daily basis
10% of the total energy intake was calculated in the form of
‘additional points’. In this regard, a list of foods with their
associated number of points was provided to the participants.
In order to ensure total energy intake and some freedom of
choice, participants were advised to choose any foods from
the list, provided they added up to the allotted number of
points for their respective energy intakes for that day. A vali-
dated FFQ and physical activity questionnaire, measuring
activity index, were analysed in order to determine the correct
energy intake requirements for the maintenance of body
weight for each participant. The validated physical activity
questionnaire is based on the Baecke physical activity ques-
tionnaire (Kruger et al. 2004). Evidence of underreporting
was found when the ratio of energy intake to BMR was less
than 1·2 (Bingham, 1991; Briefel et al. 1997). In the light of
this fact, the subjects that underreported were interviewed
again by the registered dietitian to obtain a more accurate
report on their habitual energy intake.
A 14d menu cycle was designed for five amounts of energy
intake, ranging from 8000 to 14000kJ/d (1905–3333kcal/d).
It was planned by using the FoodFinder 2 program (Medical
Research Council of South Africa, Tygerberg), which is
based on the South African food composition tables (Langen-
hoven et al. 1991). The macronutrient profiles and fatty acid
distribution of the three diets were analysed chemically to
validate the diet composition. Duplicate portions of breakfast,
lunch and dinner for the 14 d menu cycle were collected daily,
homogenized and pooled in a container and frozen at 2848C
until the analysis was done.
Compliance to dietary intervention
Quality control and compliance with the protocol were
ensured among study participants by the following means:
(1) foods were weighed to the nearest gram before being
served to the participants; (2) the principal investigator, a
registered dietitian, supervised mealtimes and ensured the
complete intake of all study foods; (3) participants kept food
diaries of the additional points used and possible left-overs
were collected and weighed (by researchers). In addition,
any deviation from the study protocol were recorded in
these diaries (these diaries were reviewed by the investigators
during the study); (4) participants were weighed twice weekly
and the energy intake was adjusted (especially during the first
3-week run-in period) in order to maintain body weight; and
(5) participants were urged to maintain the same activity
level throughout the study. Lastly, those individuals who
used chronic medication (e.g. lipid-lowering medication) at
baseline were instructed to continue use and to maintain the
same dosage for the duration of the trial.
Table 2. Macronutrient composition of 100g walnuts and unsalted cashew nuts
Walnuts Cashew nuts
tables†(g)Selected nutrientsg%‡ g %‡
Oleic acid (C18:1)
Linoleic acid (C18:2)
a-Linolenic acid (C18:3)
Palmitic acid (C16:0)
Stearic acid (C18:0)
†Langenhoven et al. (1991); Kruger et al. (1992).
‡Percentage of total fatty acids.
Blood sampling and oral glucose tolerance test
The subjects were required to fast overnight (12h). A qualified
nursing sister collected venous blood samples. For the prep-
aration of serum, 20ml blood were drawn and left to clot.
For determination of plasma glucose concentrations 5ml
blood were collected in tubes containing potassium oxalate
(10mg) and sodium fluoride (12·5mg).
After the fasting blood samples were collected, the oral glu-
cose tolerance test was continued: 75g glucose were dissolved
in 300ml water; blood samples for the measurement of glu-
cose were drawn again after 2h; blood was centrifuged for
15min at 2000g to yield serum; aliquots of serum were
stored at 2828C until the analysis was performed.
The fatty acid composition of the nuts and diets was measured
by GC as described by van Jaarsveld et al. (2000). The percen-
tage protein was analysed by a general combustion method
(AOAC Method 992·23; AOAC International, 2002) by
using a LECO FP 528 (LECO Corporation, St Joseph, MI,
USA), the percentage fat by a GAVIEZELwmethod using
the Bu ¨chi B 820 fat determination system with the Bu ¨chi B
815 extraction unit (Bu ¨chi Labortechnik AG, Flawil, Switzer-
land), the percentage fibre by the filter bag technique using the
ANKOM 220 fibre analyser with F57 filter bags (ANKOM
Technologies, Fairport, NY, USA), the percentage moisture
with the air-oven (aluminium plate) method (AACC Method
44-16l; American Association of Clinical Chemistry, 2003)
and percentage ash with the AOAC Method 942·05 (AOAC
International, 2002). The carbohydrate content was then calcu-
lated as the sum of the protein, fat, fibre, moisture and ash sub-
tracted from 100.
CV for all the laboratory analyses of blood samples were
less than 5%. Serum TC, HDL-C, TAG, uric acid and
plasma glucose were measured on a Vitros DT60 II Chemistry
System (Ortho-Clinical Diagnostics, Rochester, NY, USA)
with Vitros reagents (catalogue numbers 1532175, 1335504,
1532159, 1532134 and 1532316, respectively) and controls
(catalogue numbers 842-0317 and 144-8042). Serum LDL-C
C(mmol/l) ¼ TC 2 TAG/2·2 2 HDL).
measured using the Synchron LX20wclinical system (Beck-
man Coulter, Inc., Fullerton, CA, USA). Serum fructosamine
was measured with a calorimetric method (catalogue number
1930010; Roche, Basel, Switzerland). BP measurements
were obtained by taking a 7min continuous measurement of
cardiovascular parameters using the Finometere device
(FMS; Finapres Measurement Systems, Arnhem, Nether-
The computer software package Statisticaw(Statsoft Inc.,
Tulsa, OK, USA) was used for the analyses of the data. The
statistical analysis was done in five steps. Initially, the vari-
ables were tested for normality using the Shapiro–Wilk’s W
test. Non-normally distributed data were transformed into an
square-root transformations and again tested for normality.
Thereafter, descriptive statistics were done. Data that were
normally distributed are expressed as mean and 95% CI.
Data that are not normally distributed or logarithmic and
square-root transformed are expressed as median (25, 75 per-
centiles). Furthermore, changes within groups, from baseline
to end, were tested for significance by using the t test for
dependent samples in the case of parametric data and the Wil-
coxon matched-pairs test in the case of non-parametric data.
Also, differences in baseline and D (change from baseline to
end) between the three groups were determined by using the
ANOVA for parametricdata and the Kruskal–Wallis
ANOVA for non-parametric data. When significance between
the changes in the three groups was indicated with the
ANOVA, the Tukey honest significant difference test for
unequal N for parametric data was used to determine between
which groups the differences occurred. Lastly, weight-
adjusted differences in baseline and D between the three
groups were determined by using the analysis of covariance.
Significance was set at P#0·05.
Four subjects discontinued the study for the following reasons:
two had work obligations outside Potchefstroom, one had an
unrelated medical condition and one went on a holiday
during the study period. The remaining sixty-four subjects
(twenty-nine men and thirty-five women, aged 45 (SD 10)
years) all completed one of the three intervention diets. Com-
pliance with the experimental diets was calculated as 90%,
taking into account the left-overs, food diaries and any devi-
ation from the prescribed dietary protocols. Also, activity
level was maintained throughout the study. In this regard,
most of the participants had a sedentary lifestyle. The baseline
characteristics of the subjects did not differ between groups
(ANOVA; Table 3). Most of the subjects were obese (91%)
with high waist circumference values exceeding those indi-
cated by the ATP III criteria. Of the subjects, 53% had high
TAG concentrations, 42% had high systolic BP, 13% had
high diastolic BP, 91% had low HDL-C and only 5% had
high fasting glucose concentrations as indicated by the ATP
III criteria. At baseline 13% of subjects had high hs-CRP con-
centrations, greater than 7·5mg/l (as indicated by the produ-
characteristics at baseline were indicative of the metabolic
syndrome. Also, of the entire study population only four
were smokers (Table 3). Weight, BMI and waist circumfer-
ence remained unchanged during the intervention period.
The participants’ habitual energy intakes ranged from 5500 to
13000kJ/d (1310–3095kcal/d), however, 52% of the subjects
underreported their energy intake. The chemical analysis of
the composition of the diets was comparable to the calculated
diets, except for the total fat and carbohydrate content of the
walnut diet (Table 1). The analysed fat content was higher
because the actual fat content of the walnuts was higher
than indicated in the food composition tables (Table 2). All
three experimental diets had a low to moderate glycaemic
index and glycaemic load (Wolever & Jenkins, 1985; Foster-
Powell et al. 2002; Table 1). The composition of walnuts
(high in a-linolenic acid and linoleic acid) and cashew nuts
(high in oleic acid; Table 2) was reflected in the diets and
resulted in the anticipated increase in PUFA and MUFA con-
centrations, in the diets respectively. In particular, the fatty
acid composition of the walnut, cashew nut and control diets
(expressed as % of total fatty acids) was 6·42% a-linolenic
acid, 46·31% linoleic acid, 25·65% oleic acid; 1·05% a-lino-
lenic acid, 27·82% linoleic acid, 43·65% oleic acid and
0·86% a-linolenic acid, 27·13% linoleic acid, 32·88% oleic
In Table 4, the serum lipid concentrations displayed no signifi-
cant changes between walnut, cashew nut and control groups
at baseline and in response to the intervention. The subjects
in the control diet group showed a small significant increase
in HDL-C compared to baseline (Table 4). Results from
both nut diets displayed no significant change in HDL-C,
TAG, TC or LDL-C concentrations when compared to the
Serum fructosamine and plasma glucose
In Table 5, serum fructosamine and plasma glucose (t ¼ 0min
and t ¼ 120min) showed no significant difference at baseline
between groups. Also, there was no significant difference in
serum fructosamine between groups after the intervention
diets. Plasma glucose concentrations (t ¼ 0) increased signifi-
cantly by 0·70mmol/l (P¼0·04) in the cashew nut group com-
pared to the control group. Fructosamine significantly
increased in the control diet group at the end of the study
period, when compared to the baseline concentrations.
The 2h oral glucose tolerance test showed that there was no
significant difference between the groups and when comparing
baseline to end values.
Blood pressure, uric acid and serum high-sensitivity
Of all three intervention diets, systolic and diastolic BP as well
as uric acid concentrations displayed no significant change
between groups at baseline and from baseline to end. The
changes in S-hs CRP did not differ significantly between
groups (Table 6). There was a significant increase in S-hs
CRP concentrations in the walnut intervention group from
baseline to end and no significant change in the cashew nut
and control groups when comparing baseline to end values.
Although not statistically significant, S-hs CRP was also
increased, approximately to the same extent, in the control
and cashew nut groups. The increase in the walnut group is,
therefore, probably not an independent effect of walnuts.
As far as we know this well-designed parallel, randomized,
controlled feeding trial investigating the effects of nuts on par-
ticipants having the metabolic syndrome is the first study of its
kind. Regarding our main objective, we found that both the
walnut and the unsalted cashew nut intervention diets had
no significant effect on the HDL-C, TAG, TC, LDL-C,
serum fructosamine, S-hs CRP, BP and serum uric acid con-
centrations when compared to the control diet. Plasma glucose
concentrations increased significantly in the cashew nut group
compared to the control group.
In a recent systematic review (Mukuddem-Petersen et al.
2005), it was concluded from randomized controlled interven-
tion trials that the consumption of 50–100g (approximately
1·5–3·5 servings) of nuts five or more times/week as part of
a heart healthy diet with a total fat content (high in MUFA
and/or PUFA) of approximately 35%E may significantly
decrease TC and LDL-C. In lieu of the dynamic make-up of
nuts, it deserves a mention that this decrease is not solely
due to the changes in the fatty acid composition that results
when nuts are included in the diet, but also as a result of
the other components found in nuts.
Contrary to the outcomes of the present controlled feeding
trial, four out of seven well-designed walnut studies (40–
84g/d) displayed a significant decrease in TC and LDL-C
when compared to Step I (healthy; Sabate et al. 1993), Med-
iterranean (hypercholesterolaemic; Zambon et al. 2000; Ros
et al. 2004) and Japanese (healthy; Iwamoto et al. 2002)
diets. The average %E from fat in the aforementioned nut
diets and the control diets were 31 and 29%, respectively.
Two of the studies that showed no significant change in the
lipid profile of hyperlipidaemic subjects who followed a
walnut intervention diet (64–78g/d) when compared to a
low-fat (30%E from fat) and Step I diet (33%E from fat) pro-
vided $38%E from fat (Chisholm et al. 1998; Morgan et al.
2002), suggesting that the beneficial effects of nuts disappear
with high fat intakes. Similarly, in the current study, the
walnut diet (60–100g/d) was high in fat (41%E from fat)
Table 3. Characteristics at baseline for all subjects
Walnut diet (n 21)Cashew nut diet (n 21)Control diet (n 22)
VariablesMean 95% CIMean 95% CI Mean95% CI
WC, waist circumference.
Table 4. Weight and serum lipids during interventions*
Walnut diet (n 21)
(B v. E)
Cashew nut diet (n 21)
(B v. E)
Control diet (n 22)
(B v. E)
P value between
groups (ANOVA) Mean 95% CIMean 95% CI Mean 95% CI
2 0·87, 0·44
2 1·13, 0·79
2 1·19, 0·17
2 0·10, 0·04
2 0·35, 0·27
2 0·09, 0·06
2 0·52, 0·20
2 0·16, 0·38
0·96TAG 0·790·36 0·41
(B v. E) Median
(B v. E) Median
(B v. E)
P value between
2 0·30, 0·37
2 0·21, 0·50
2 0·22, 0·49
2 0·25, 0·55
2 0·06, 0·60
2 0·11, 0·46
B, baseline (after 3-week run-in period); B v. E, P values for change from baseline to end of intervention period; E, end; HDL-C, HDL cholesterol; LDL-C, LDL cholesterol; TC, total cholesterol; D, change from baseline to end.
*Significance was set at P#0·05.
Table 5. Serum fructosamine and plasma glucose during interventions
(B v. E)
Cashew nut diet
(B v. E)
(B v. E)
P value between
groups (ANOVA)MarkersMean 95% CIMean 95% CIMean95% CI
B210202, 2180·79 212206, 2180·32 204198, 2100·040·23
2 8·16, 6·26
2 2·38, 6·96
0·20, 11·8 2·29 6·000·24
(B v. E)Median
(B v. E) Median
(B v. E)
P value between
Plasma glucose t¼0
B4·50 4·30, 5·200·644·70 4·30, 5·100·034·554·30, 5·400·18 0·73
2 0·60, 0·70
2 0·30, 1·60
2 1·40, 0·80
0·94Plasma glucose. t ¼ 120
0·09 0·25 0·87
2 0·20, 1·30
2 0·90, 1·100·80
B, baseline (after 3-week run-in period); B v. E, P values for change from baseline to end of intervention period; E, end; D, change from baseline to end.
Median value was significantly different from that of the control group: *P#0·05.
Table 6. Blood pressure, uric acid and C-reactive protein changes within and between groups during intervention*
(B v. E)
Cashew nut diet
(B v. E)
(B v. E)
P value between
groups (ANOVA)Markers Mean 95% CI Mean95% CIMean 95% CI
2 2·39, 4·81
2 7·09, 2·19
2 4·01, 2·83
2 2·45, 5·92
2 3·11, 4·07
2 12·2, 37·3
2 13·5, 21·2
2 12·1, 21·43·864·640·78
(B v. E)Median
(B v. E) Median
(B v. E)
P value between
S-hs CRP (mg/l)B
2 0·30, 2·20
B, baseline; B v. E, P values for change from baseline to end of intervention period; BP, blood pressure; E, end; S-hs CRP, serum high-sensitivity C-reactive protein; D, change from baseline to end.
*Significance was set at P#0·05.
compared to the control diet (33%E from fat). However, high-
PUFA diets, similar to the walnut diet in the current study, are
predicted to reduce TC and LDL-C concentrations (Mensink
& Katan, 1992). Based on the predictive equations of Mensink
& Katan (1992), which predict changes in TC and LDL-C
concentrations according to changes in the fatty acid content
of the diet, the walnut diet should have reduced the TC and
LDL-C concentrations with 0·34 and 0·30mmol/l, respect-
ively, compared to the control diet. The lack of effect can
possibly be ascribed to the low TC and LDL-C concentrations
of the subjects at baseline, as these low concentrations make it
difficult to show a further benefit.
Most of the walnut studies to date did not show an effect on
HDL-C and TAG concentrations (Mukuddem-Petersen et al.
2005). Sabate et al. (1993) showed a decrease in HDL-C
that could possibly have been due to the high PUFA content
on the walnut diet (17%E) compared to the control diet
(10%E). High intakes of PUFA (.10%E) may decrease
HDL-C (Riccardi et al. 2003). Although the PUFA content
of the walnut diet in the current study was also much higher
compared to the control diet (21·4 v. 9·5%E), the HDL-C
did not differ between the diets.
As no clinical trials have been done on cashew nuts before,
we considered almond studies as they are similar in compo-
sition to cashew nuts. We expected the cashew nut study to
have a beneficial effect on TC and LDL-C concentrations as
seen in previous almond nut studies. In this context three
out of four well-designed almond studies (54–100g/d) ran-
ging from 32 to 39%E from fat, significantly decreased TC
and LDL-C in hypercholesterolaemic (Spiller et al. 1998; Jen-
kins et al. 2002) and normocholesterolaemic (Sabate et al.
2003) subjects compared to subjects on a control diet
(35%E from fat), low-fat (26·3%E from fat) and Step I diet
(30%E from fat) (Sabate et al. 2003). In the present study,
the cashew nut (66–115g/d) diet (37%E from fat) had no sig-
nificant beneficial effect on the lipid profile (Table 4) when
compared to the control diet that is lower in fat (33%E
from fat). This outcome was very similar to results found by
Lovejoy et al. (2002) who showed that 57–113g almonds/d
(39%E from fat) had no significant beneficial effect on the
lipid profile of diabetic subjects when compared to a high-
fat (37%E from fat) control group. The same outcome was
achieved when the same amount of almonds (27%E) as part
of a low-fat diet was compared to a low-fat (26%E from
fat) control diet (Lovejoy et al. 2002). Based on differences
in fatty acid content of the cashew nut and control diets in
the current study the predicted reductions in TC and LDL-C
concentrations with the cashew nut diet were very small
(20·14mmol/l for both TC and LDL-C). However, it has
been shown that nuts may reduce TC and LDL-C concen-
trations beyond the effects predicted based solely on fatty
acid profiles (Mukuddem-Petersen et al. 2005). As mentioned
earlier, the low TC and LDL-C concentrations at baseline
might explain the lack of effects. Other possible reasons for
the non-significance in lipid concentrations need to be
No direct studies have been done to investigate the effects
of nuts on BP. Despite this, we anticipated an improvement
in the BP readings with the walnut and cashew nut diets
based on their fatty acid composition. In particular, emerging
research has suggested possible health benefits associated with
modest increases in dietary a-linolenic acid (walnuts),
including reduced BP (Ferrara et al. 2000; Hermansen,
2000; Hunter, 1990). Also, numerous studies conducted in
healthy and hypertensive individuals have shown a beneficial
effect of MUFA (cashew nuts) on a number of outcomes
related to cardiovascular risk, including BP (Roche et al.
1998). However, no improvement in BP readings was seen
after the nut intervention diets in the current controlled feed-
Zhao et al. (2004) concluded that a diet high in a-linolenic
acid, obtained from walnuts, walnut oil and flaxseed oil, eli-
cited cardioprotective effects and vascular anti-inflammatory
effects. Regarding the latter, it has been reported that walnuts
are amongst the dietary plants that contain the most antioxi-
dants (Halvorsen et al. 2002) and several health effects have
been ascribed to flavonoids (antioxidants) including reduced
inflammation (Nijveldt et al. 2001). In the present study, the
walnut diet (high in a-linolenic acid) resulted in an increase
in S-hs CRP concentrations, although this was probably not
an independent effect of walnuts. Recently, some evidence
has been presented for a beneficial effect of MUFA on a
number of outcomes related to cardiovascular risk, including
reduced inflammation (Ferrara et al. 2000). Consequently, a
cashew nut diet (high in MUFA) could be expected to improve
the inflammatory parameter CRP. However, this was not evi-
dent in the present study.
Numerous studies compared a low-fat diet (21%; 23%;
29%E from fat; Luscombe et al. 1999; Thomsen et al.
1999; Rodriguez-Villar et al. 2000; Perez-Jimenez et al.
2001) to a high-MUFA diet (35%; 40%E from fat; Luscombe
et al. 1999). The results of these studies provided similar gly-
caemic control. Those authors concluded that provided the
intake of SFA is low, a MUFA diet with a total fat content
of up to 40%E has effects on glycaemic control that are simi-
lar to those of the traditional high-carbohydrate diet with fat
limited to 25–30%E. An almond nut study conducted by
Lovejoy et al. (2002) showed no effect on plasma glucose
concentrations compared to a low-fat (26%E from fat) control
group (olive and rapeseed oil) in diabetics. In contrast to these
previous findings, the significant increase of the plasma blood
glucose (t ¼ 0) seen in the cashew nut diet (high MUFA;
36·5%E from fat) group was unexpected. Even though
serum fructosamine is a short-term marker of glycaemic con-
trol, its concentrations remained unchanged in the cashew nut
group compared to the control group that increased. It could
be speculated that the aforementioned increase in serum fruc-
tosamine concentrations in the control group may be due to
this group having lower baseline values compared to the
other two groups.
ATP III recommends that obesity should be the primary
target of intervention for the metabolic syndrome. In turn,
the first line of therapy should be weight reduction reinforced
with increased physical activity. A notion related to this is evi-
dent in the study by Esposito et al. (2004) where they dis-
played how a Mediterranean diet (high in MUFA and
PUFA) that included nuts, weight loss and increased physical
activity resulted in a significant reduction in S-hs CRP
amongst other beneficial anti-inflammatory responses.
In conclusion, individuals having the metabolic syndrome
showed no improvement in the markers of this syndrome
after following a walnut diet or a cashew nut diet (8 weeks)
compared to a control diet while maintaining body weight.
Most of the study population was obese (average BMI 35)
and sedentary therefore it could be speculated that with such
a high degree of obesity that even a ‘good diet including
nuts’ would not suffice in inducing beneficial effects without
weight loss. Firstly, it may be suggested that maintenance of
body weight may have masked the positive metabolic effects
of the nut diets. Especially, if these diets mediate its beneficial
effects mainly through central appetite suppression and conse-
quent body weight reduction. Wien et al. (2003) showed how
a low-energy nut diet resulted in sustained weight reduction
and improved the preponderance of abnormalities associated
with the metabolic syndrome. Future nut research on these
aspects are worthy of exploration.
A. Greyling, C. de Witt, J. Wheeler, J. Bekker, C. Jansen van
Rensburg, M. Bailey, L. Loots, V. van Scheltinga, L. Wiggett,
L. Davies, L. van Wyk, E. Snyman, F. Mpho and T. Holele, A.
Schutte, H. Huisman, J. van Rooyen, S. Jordaan, F. van der
Westhuizen, C. S. Venter, H. H. Wright, H. S. Kruger, H.
van’t Riet, R. Breet, A. Van Graan, M. van Lieshout, D. Loots,
M. Pieters-Loots, K. Moruisi, M. Opperman and M. Phometsi
for contributing in various ways to the successful execution
and completion of the controlled feeding trial – The Nut
Study. Financial support was received from The National
Research Foundation (NRF) and the South African govern-
were received from the following food companies: Tiger
Brands, Pick ’n Pay, Clover and Unilever. We declare that we
have no conflict of interest.
American Association of Clinical Chemistry (2003) Abstracts of the
55th Annual Meeting of the AACC, Philadelphia, Pennsylvania,
USA, July 20–24. Clin Chem 20 (49), A2–A182.
AOAC International (2002) Official Methods of Analysis of AOAC
International. Gaithersburg, MD: AOAC International.
Bingham SA (1991) Limitations of the various methods for collecting
dietary intake data. Ann Nutr Metab 35, 117–127.
Briefel RR, Sempos CT, McDowell MA, Chien S & Alaimo K (1997)
Dietary methods research in the third National Health and Nutrition
Examination Survey: underreporting of energy intake. Am J Clin
Nutr 65, 1203S–1209S.
Chisholm A, Mann J, Skeaff M, Frampton C, Sutherland W, Duncan A
& Tiszavari S (1998) A diet rich in walnuts favourably influences
plasma fatty acid profile in moderately hyperlipidaemic subjects.
Eur J Clin Nutr 52, 12–16.
Dansinger ML & Schaefer EJ (2006) Low-carbohydrate or low-fat
diets for the metabolic syndrome? Curr Diabet Rep 6, 55–63.
Dreher ML, Maher CV & Kearney P (1996) The traditional and emer-
ging role of nuts in healthful diets. Nutr Rev 54, 241–245.
Esposito K, Marfella R, Ciotola M, et al. (2004) Effect of a mediter-
ranean-style diet on endothelial dysfunction and markers of vascu-
lar inflammation in the metabolic syndrome: a randomized trial.
JAMA 292, 1440–1446.
Ferrara LA, Raimondi AS, D’Episcopo L, Guida L, Dello RA &
Marotta T (2000) Olive oil and reduced need for antihypertensive
medications. Arch Intern Med 160, 837–842.
Ford ES (2004) Prevalence of the metabolic syndrome in US popu-
lations. Endocrinol Metab Clin N Am 33, 333–350.
Ford ES, Giles WH & Dietz WH (2002) Prevalence of the metabolic
syndrome among US adults: findings from the third National
Health and Nutrition Examination Survey. JAMA 287, 356–359.
Foster-Powell K, Holt SH & Brand-Miller JC (2002) International
table of glycemic index and glycemic load values: 2002. Am J
Clin Nutr 76, 5–56.
Freire RD, Cardoso MA, Gimeno SGA & Ferreira SRG (2005)
Dietary fat is associated with metabolic syndrome in Japanese
Brazilians. Diabetes Care 28, 1779–1785.
Fujioka K, Greenway F, Sheard J & Ying Y (2006) The effects of
grapefruit on weight and insulin resistance: relationship to the
metabolic syndrome. J Med Food 9, 49–54.
Giacco R, Parillo M, Rivellese AA, et al. (2000) Long-term dietary
treatment with increased amounts of fiber-rich low-glycemic
index natural foods improves blood glucose control and reduces
the number of hypoglycemic events in type 1 diabetic patients.
Diabetes Care 23, 1461–1466.
ing of total antioxidants in dietary plants. J Nutr 132, 461–471.
He K, Liu K, Daviglus ML, Morris SJ, Loria CM, Van Horn L,
Jacobs DR Jr & Savage PJ (2006) Magnesium intake and incidence
of metabolic syndrome among young adults. Circulation 113,
Hermansen K (2000) Diet, blood pressure and hypertension. Br J Nutr
83 (Suppl. 1), S113–S119.
Hunter JE (1990) n-3 fatty acids from vegetable oils. Am J Clin Nutr
Iwamoto M, Imaizumi K, Sato M, et al. (2002) Serum lipid profiles in
Japanese women and men during consumption of walnuts. Eur J
Clin Nutr 56, 629–637.
Jenkins DJ, Kendall CW, Marchie A, et al. (2002) Dose response of
almonds on coronary heart disease risk factors: blood lipids,
oxidized low-density lipoproteins, lipoprotein(a), homocysteine,
and pulmonary nitric oxide: a randomized, controlled, crossover
trial. Circulation 106, 1327–1332.
Jimenez-Cruz A, Seimandi-Mora H & Bacardi-Gascon M (2003)
Effect of low glycemic index diets in hyperlipidemia. Nutr Hosp
Kruger HS, Venter CS & Steyn HS (2004) A standardized physical
activity questionnaire for a population in transition: the Thusa
study. Afr J Phys Health Educ Recreation Dance 6, 54–64.
Composition Tables. Supplement to MRC Food Composition Tables
(1991). Tygerberg: Medical Research Council of South Africa.
Langenhoven M, Kruger M, Gouws E & Faber M (1991) MRC Food
Composition Tables. Supplement, 3rd ed., Tygerberg: Medical
Research Council of South Africa.
Lovejoy JC, Most MM, Lefevre M, Greenway FL & Rood JC (2002)
Effect of diets enriched in almonds on insulin action and serum
lipids in adults with normal glucose tolerance or type 2 diabetes.
Am J Clin Nutr 76, 1000–1006.
Luscombe ND, Noakes M & Clifton PM (1999) Diets high and low in
glycemic index versus high monounsaturated fat diets: effects on
glucose and lipid metabolism in NIDDM. Eur J Clin Nutr 53,
McKeown NM, Meigs JB, Liu S, Saltzman E, Wilson PW & Jacques
PF (2004) Carbohydrate nutrition, insulin resistance, and the preva-
lence of the metabolic syndrome in the Framingham Offspring
Cohort. Diabetes Care 27, 538–546.
Mensink RP & Katan MB (1992) Effect of dietary fatty acids on
serum lipids and lipoproteins. A meta-analysis of 27 trials. Arter-
ioscler Thromb 12, 911–919.
Michalsen A, Lehmann N, Pithan C, Knoblauch NTM, Moebus S,
Kannenberg F, Binder L, Budde T & Dobos GJ (2006) Mediterra-
nean diet has no effect on markers of inflammation and metabolic
risk factors in patients with coronary artery disease. Eur J Clin
Nutr 60, 478–485.
Morgan JM, Horton K, Reese D, Carey C, Walker K & Capuzzi DM
(2002) Effects of walnut consumption as part of a low-fat, low-
cholesterol diet on serum cardiovascular risk factors. Int J Vitam
Nutr Res 72, 341–347.
Moruisi KG, Oosthuizen W & Opperman AM (2006) Phytosterols/
stanols lower cholesterol concentrations in familial hypercholester-
olemic subjects: a systematic review with meta-analysis. J Am Coll
Nutr 25, 41–48.
Mukuddem-Petersen J, Oosthuizen W & Jerling JCJ (2005) A sys-
tematic review of the effects of nuts on blood lipid profiles in
humans. J Nutr 135, 2082–2089.
Muzio F, Mondazzi L, Sommariva D & Brachi A (2005) Long-term
effects of low-calorie diet on the metabolic syndrome in obese non-
diabetic patients. Diabetes Care 28, 1485–1486.
National Cholesterol Education Program (2001) Executive summary
of the third report of the National Cholesterol Education Program
(NCEP) Expert Panel on Detection, Evaluation, and Treatment of
High Blood Cholesterol in Adults (Adult Treatment Panel III).
JAMA 285, 2486–2497.
Nijveldt RJ, van Nood E, van Hoorn DE, Boelens PG, van Norren K
& van Leeuwen PA (2001) Flavonoids: a review of probable mech-
anisms of action and potential applications. Am J Clin Nutr 74,
Oosthuizen W, Vorster HH, Kruger A, Venter CS, Kruger HS &
de Ridder JH (2002) Impact of urbanisation on serum lipid profiles
– the THUSA survey. S Afr Med J 92, 723–728.
Montilla P, Marin C, Velasco MJ, Blanco-Molina A, Jimenez
Pereperez JA & Ordovas JM (2001) A Mediterranean and a high-
carbohydrate diet improve glucose metabolism in healthy young
persons. Diabetologia 44, 2038–2043.
Pieters M, Oosthuizen W, Jerling JC, Loots DT, Mukuddem-Petersen
J & Hanekom SM (2005) Clustering of haemostatic variables and
the effect of high cashew and walnut diets on these variables in
metabolic syndrome patients. Blood Coag Fibrin 16, 429–437.
Riccardi G, Giacco R & Rivellese AA (2004) Dietary fat, insulin sen-
sitivity and the metabolic syndrome. Clin Nutr 23, 447–456.
Riccardi G, Rivellese AA & Williams CM (2003) The cardiovascular
system. In Nutrition and Metabolism, pp. 224–246 [MJ Gibney, IA
Macdonald and HM Roche, editors]. Oxford: Blackwell Science.
Roche HM, Zampelas A, Knapper JM, et al. (1998) Effect of long-
term olive oil dietary intervention on postprandial triacylglycerol
and factor VII metabolism. Am J Clin Nutr 68, 552–560.
Rodriguez-Villar C, Manzanares JM, Casals E, Perez-Heras A,
Zambon D, Gomis R & Ros E (2000) High-monounsaturated fat,
olive oil-rich diet has effects similar to a high-carbohydrate diet
on fasting and postprandial state and metabolic profiles of patients
with type 2 diabetes. Metabolism 49, 1511–1517.
Ros E, Nunez I, Perez-Heras A, et al. (2004) A walnut diet improves
endothelial function in hypercholesterolemic subjects: a randomized
crossover trial. Circulation 109, 1609–1614.
Sabate J (1993) Does nut consumption protect against ischaemic heart
disease? Eur J Clin Nutr 47, Suppl. 1, S71–S75.
Sabate 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. N Engl J Med 328, 603–607.
Sabate J, Haddad E, Tanzman JS, Jambazian P & Rajaram S (2003)
Serum lipid response to the graduated enrichment of a Step I diet
with almonds: a randomized feeding trial. Am J Clin Nutr 77,
Shirai K (2004) Obesity as the core of the metabolic syndrome and
the management of coronary heart disease. Curr Med Res Opin
Spiller GA, Jenkins DA, Bosello O, Gates JE, Cragen LN & Bruce B
(1998) Nuts and plasma lipids: an almond-based diet lowers LDL-
C while preserving HDL-C. J Am Coll Nutr 17, 285–290.
Thomsen C, Rasmussen O, Christiansen C, Pedersen E, Vesterlund M,
of a monounsaturated fat diet and a high carbohydrate diet on cardio-
Eur J Clin Nutr 53, 818–823.
Tapsell LC, Gillen LJ, Patch CS, Batterham M, Owen A, Bare M &
Kennedy M (2004) Including walnuts in a low-fat/modified-fat diet
improves HDL cholesterol-to-total cholesterol ratios in patients
with type 2 diabetes. Diabetes Care 27, 2777–2783.
van Jaarsveld PJ, Smuts CM, Tichelaar HY, Kruger M & Benade AJ
(2000) Effect of palm oil on plasma lipoprotein concentrations and
primates. Int J Food Sci Nutr 51, Suppl., S21–S30.
Wien MA, Sabate JM, Ikle’ DN, Cole SE & Kandeel FR (2003)
Almonds vs complex carbohydrates in a weight reduction program.
Int J Obes 27, 1365–1372.
Wirfalt E, Hedblad B, Gullberg B, et al. (2001) Food patterns and
components of the metabolic syndrome in men and women:
a cross-sectional study within the Malmo Diet and Cancer
cohort. Am J Epidemiol 154, 1150–1159.
Wolever TM & Jenkins DJ (1985) Application of glycaemic index to
mixed meals. Lancet 2, 944.
Zambon D, Sabate J, Munoz S, et al. (2000) Substituting walnuts for
monounsaturated fat improves the serum lipid profile of hypercho-
lesterolemic men and women. A randomized crossover trial. Ann
Intern Med 132, 538–546.
Zhao G, Etherton TD, Martin KR, West SG, Gillies PJ &
Kris-Etherton PM (2004) Dietary a-linolenic acid reduces inflam-
matory and lipid cardiovascular risk factors in hypercholesterole-
mic men and women. J Nutr 134, 2991–2997.
Zibaeenezhad MJ, Shamsnia SJ & Khorasani M (2005) Walnut con-
sumption in hyperlipidemic patients. Angio 56, 581–583.