Atherosclerosis 211 (2010) 630637

Contents lists available at ScienceDirect

Atherosclerosis
journal homepage: www.elsevier.com/locate/atherosclerosis

Soluble bre (Plantago ovata husk) reduces plasma low-density lipoprotein (LDL)
cholesterol, triglycerides, insulin, oxidised LDL and systolic blood pressure in
hypercholesterolaemic patients: A randomised trial
Rosa Sol a, , Eric Bruckert b , Rosa-Maria Valls a , Silvia Narejos c , Xavier Luque d , Manuel Castro-Cabezas e ,
Gema Domnech f , Ferran Torres f , Mercedes Heras a , Xavier Farrs g , Jos-Vicente Vaquer h ,
Jos-Miguel Martnez i , Maria-Cruz Almaraz j , Anna Anguera k
a

Unitat de Recerca en Lpids i Arteriosclerosi, CIBERDEM, Hospital Universitari Sant Joan, IISPV, Universitat ROVIRA i VIRGILI, Reus, Spain
Hpital Piti-Salptrire, Service dEndocrinologie-mtabolisme, Paris, France
c
CAP Centelles, Centelles, Spain
d
CAP Alcover, Alcover, Spain
e
Hospital Sant Franciscus Gasthuis, Rotterdam, The Netherlands
f
Laboratory of Biostatistics & Epidemiology (Universitat Autonoma de Barcelona), Statistics & Methodology Support Unit (USEM), IDIBAPS, (Hospital Clinic), Barcelona, Spain
g
CAP El Remei, Vic, Spain
h
Centro de Salut Petrer I, Petrer, Spain
i
Centro de Salut Toms Ortu
no, Benidorm, Spain
j
Hospital Civil Pabelln C, Servicio de Endocrinologa y Nutricin, Mlaga, Spain
k
Rottapharm|Madaus, S.L., Barcelona, Spain
b

a r t i c l e

i n f o

Article history:
Received 6 November 2009
Received in revised form 5 March 2010
Accepted 7 March 2010
Available online 17 March 2010
Keywords:
Soluble bre
Cholesterol LDL
Blood pressure
Insulin
Oxidised LDL

a b s t r a c t
The objective was to evaluate whether the soluble bre Plantago ovata (Po)-husk improves cardiovascular
disease (CVD) risk biomarkers including low-density lipoprotein cholesterol (LDL-C).
Methods: In a multi-centred, double-blind, placebo-controlled, parallel, randomised trial conducted
in primary care-clinics in Spain, France and Holland, mild-moderate hypercholesterolaemic patients
(age range: 4367 years) received 14 g/d of Po-husk (n = 126) or placebo (microcrystalline-cellulose
14 g/d; n = 128) in a low saturated fat diet for 8 weeks. Subsequently, if LDL-C remained 3.35 mmol/L
[130 mg/dL], participants proceeded with the bre plus simvastatin (20 mg/d) for further 8 weeks. Lipid
prole, blood pressure (BP), insulin, oxidised LDL and some gene polymorphisms involved in CVD risk
were measured.
Results: Relative to placebo, Po-husk reduced plasma LDL-C by 6% (P < 0.0002), total cholesterol (TC)
by 6%, triglycerides (TG) by 21.6%, apolipoprotein (Apo) B-100 by 6.7%, oxidised LDL by a mean of
6.82 U/L (95%CI: 3.1510.48), insulin by 4.68 pmol/L (95%CI: 0.688.67) and systolic BP by 4.0 mm
Hg (95%CI; 1.26.7) (P < 0.05). The TG-lowering effect in the Po-husk group was magnied by variants in
plasminogen-activator-inhibitor (PAI-1; rs1799768) and fatty acid-binding protein (FABP-2; rs1799883)
genes. At 16 weeks, the intra-group action of simvastatin (20 mg/d) added to Po-husk or placebo was a
similar LDL-C reduction.
Conclusions: Po-husk, apart from lowering LDL-C, also reduced TG, TG related to certain gene variants,
TC, Apo B-100, oxLDL, insulin-resistance and systolic BP in mild-moderate hypercholesterolaemic
individuals.
Thus, the target patients to receive Po-husk would be those who present a cluster of various CVD risk
factors, such as metabolic syndrome.
2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Corresponding author at: Unitat de Recerca en Lpids i Arteriosclerosi, Facultat

de Medicina i Cincies de la Salut, C/Sant Llorenc 21, 43201 Reus, Spain.
Tel.: +34 977 759 369; fax: +34 977 759 322.
E-mail address: rosa.sola@urv.cat (R. Sol).
0021-9150/$ see front matter 2010 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.atherosclerosis.2010.03.010

Increased dietary bre intakes are associated with signicantly

lower prevalence rates of cardiovascular disease (CVD) [13]. A
debate is whether bre supplements to the diet provide additional CVD benets [14]. Soluble bre, when included within a low
saturated fat and cholesterol diet, lowers low-density lipoprotein

R. Sol et al. / Atherosclerosis 211 (2010) 630637

cholesterol (LDL-C) concentration about 510% in hypercholesterolaemic and diabetic patients [35]. Depending on specic
characteristics, a soluble bre such as Plantago ovata (Po) husk has
been shown to reduce triglyceridaemia in human secondary CVD
risk trials [6] and in obese Zucker rats [7]. Effects on CVD risk factors
such as lowering arterial hypertension [3], improving control of diabetes mellitus [3,4,8] and decreasing excess bodyweight [9] have
been described with other soluble bre types in human trials [39].
The recommendations for any individual with a moderately high
risk of CVD due to lifestyle-related risk factors (such as obesity,
physical inactivity) and elevated triglycerides (TG), low highdensity lipoprotein cholesterol (HDL-C), or metabolic syndrome are
to modify these risk factors regardless of LDL-C levels. However, if
LDL-C concentration remains 3.35 mmol/L [130 mg/dL] following
these therapeutic life-style changes, a LDL-lowering agent needs to
be prescribed to achieve a goal of <3.35 mmol/L [130 mg/dL] [1,2].
Differences in genetic background affect the response to diet
[10], but soluble bre addition to the diet has yet to be evaluated.
Our hypothesis is that the soluble bre P. ovata husk (Po-husk)
lowers LDL-C plasma concentrations and, as well, could have benecial effects on other biomarkers of CVD risk. We chose 14 g of
soluble bre based on the results obtained from the only published
study assessing therapeutic hypcholesterolaemic effects of Po-husk
in type 2 diabetic patients [4]. This study had shown that an intake
of 14 g/d Po-husk was well tolerated by the participants, and caused
a reduction of 7% of total cholesterol (TC) and 9% of LDL-C [4].
The aim was to assess the efcacy of Po-husk (14 g/d for 8
weeks) in reducing plasma LDL-C by at least 5%. The effect on
other CVD factors such as TG, apolipoprotein (Apo), oxidised LDL
(oxLDL), insulin, antithrombotic and inammation biomarker concentrations in plasma, together with insulin-resistance and blood
pressure (BP) were to be assessed. If, after Po-husk treatment or
placebo for 8 weeks, LDL-C remained 3.35 mmol/L [130 mg/dL],
a cholesterol-lowering statin (simvastatin) was to be added for a
further 8 weeks. Response to Po-husk treatment associated with
some DNA polymorphisms involved in lipid metabolism and CVD
risk were studied.

631

Spain, one hospital in France and one hospital in The Netherlands.

The Hospital Universitari Sant Joan de Reus (Spain) was the coordinating centre.
All the investigators used the same standardised protocol,
written manuals, specic guidelines, and materials used to train
health-care personnel to uniformly deliver the intervention. The
protocol was approved by the appropriate Clinical Research Ethics
Committees. The protocol was explained in a training session for all
the investigators, and the coordinating centre attended-to all questions and problems arising in the course of the study follow-up. All
participants provided written informed consent prior to enrolment
into the trial.
The participants were community-dwelling men and women
>20 years of age, with LDL-C 3.35 mmol/L [130 mg/dL] and
4.88 mmol/L [189 mg/dL], and at least one other major CVD risk
factor (age >45 years in men and >55 in women, smoking habit,
high BP, HDL-C 1.03 mmol/L [40 mg/dL] in men and 1.19 mmol/L
[46 mg/dL] in women, or family history of premature heart disease). Exclusion criteria included diabetes mellitus, any chronic
disease, hypolipaemic treatment, TG >3.97 mmol/L [350 mg/dL],
BMI >35 kg/m2 , and a history of cardiovascular disease.
Participant eligibility or exclusion was assessed by the attending
physician and was based on review of clinical records, followed by
a screening visit.
2.3. Randomisation and intervention

The study was a phase IV-II, multi-centred, randomised, doubleblind, placebo-controlled, two-arm study of the effect of Po-husk
on hypercholesterolaemia.
The primary endpoint was the reduction (between baseline
and week-8) of LDL-C concentration by 5% as a result of Po-husk
treatment. Secondary endpoints were the change from baseline to
week-8 in plasma lipids, ApoA-1, ApoB-100, inammation markers,
oxLDL, insulin-resistance and BP. Subsequently, at week-8, those
individuals who had LDL-C levels 3.35 mmol/L [130 mg/dL], were
prescribed simvastatin (20 mg/d), and the effects of the combined
treatments were assessed at week-16. Po-husk treatment response
was assessed in relation to DNA polymorphisms of candidate genes.
Adverse events were coded according to the MedDra dictionary
(version 8.0).
The study was approved by the Clinical Research Ethical Committees of all participating centres as well as the AEMPS (Spanish
Medicines Agency). Protocols were according to the Helsinki Declaration. This trial was registered with ClinicalTrials.gov: number
NCT00502047 and EudraCT No. year-2004-002184-24.

The randomisation code was computer generated. Participant

assignment to treatment or placebo arm was at a ratio of 1:1. The
number sequence for the subject, centre, and treatment assignment
were allocated via an interactive electronic response system, by the
Barcelona Randomization Unit, which took no further part in the
study.
After a 2-week run-in period, screened patients were randomised to receive Po-husk (Plantaben , Madaus S.A., Barcelona,
Spain) treatment (14 g/d) or placebo (14 g/d) for 8 weeks. Blinding
was maintained using matching placebo sachets which did not differ from the active bre with respect to colour, appearance, or any
other physical characteristic.
Po-husk was manufactured as a palatable, orange-avoured,
sugar-free product, distributed in 4 sachets of 5 g each (70% soluble bre). The dosage was one sachet dissolved in 200 mL of water,
taken 15 min before breakfast and lunch and two before the evening
meal. A 20 mg/d simvastatin tablet (Cinfa Laboratorios, Spain) was
taken 2 h after the last meal at night.
Treatment monitoring of compliance was by counting the
unopened sachets and/or simvastatin blisters returned at follow-up
visits. Consumption >80% was considered acceptable.
During the run-in period, the percentage of saturated fatty
acids (SFA) in the diet was 1013% and in the intervention period,
the SFA was set at 7% within an isocaloric diet calculated using
the HarrisBenedict equation [1,2]. At each of four clinical visits, diet compliance was monitored using 3-d dietary records and
interviews with the dietician. On occasions the participants were
contacted by telephone without warning and invited to ll-in a 24h dietary recall. In each visit, standard anthropometric data were
obtained while participants were wearing lightweight clothing and
no shoes. The data were collected by trained study personnel using
a calibrated balance, and, if the weight varied by >1 kg, the energy
intake was modied. All participants were advised to maintain their
usual physical activity.

2.2. Participants and recruitment

2.4. Measurements

Between September 2005 and June 2007, eligible outpatients

were recruited in ve primary care centres and two hospitals in

Fasting blood sample was taken at 0 and at weeks 8 and 16.

With the subject seated, BP was measured three times at 1-min

2. Subjects and methods

2.1. Design

632

R. Sol et al. / Atherosclerosis 211 (2010) 630637

Fig. 1. Flow of subjects through the study. Note: *There were patients excluded for more than one reason.

intervals using an automatic sphygmomanometer (OMRON HEM907; Peroxfarma, Barcelona, Spain) and the mean value recorded.
Screening chemistries and haemogram were performed in
each of the participating centres. All lipids and other biomarkers were measure centrally at Hospital Universitari Sant Joan, Reus
(Catalunya, Spain). Samples were stored at 80 C in the central
laboratorys Biobanc (bancmb@grupsagessa.com). TC, TG, HDL-C,
ApoA-1, ApoB-100, high sensitivity C-reactive protein (hs-CRP) and
glucose measurements were performed using standard methods
on an autoanalyzer (Beckman Coulter-Synchron, Galway, Ireland).
LDL-C was calculated by means of the Friedewald formula [11].
Interleukin 6 (IL-6), ultrasensitive insulin and folic acid were
measured by autoanalyzer (Beckman Coulter-Access Immunoassay Systems, Galway, Ireland). ELISA kits were used to measure
plasma oxLDL (Mercodia AB, Uppsala, Sweden), plasminogenactivator-inhibitor 1 (PAI-1; Technoclone GmbH, Vienna, Austria),
vascular adhesion molecule-1 (VCAM)-1 and intercellular adhesion molecule-1 (ICAM)-1 (R&D Systems, Minneapolis, USA) and
the HOMA-index was calculated [12].
DNA from each subject was obtained by the cell package
commercial kit (Servicios Hospitalarios, Spain) and used for 14
gene polymorphism measurements as single nucleotide polymor-

phisms (SNP). These were: Apo E (APOE; APOE-112: apolipoprotein

E rs7412; APOE-158: apolipoprotein E rs429358), angiotensin
converting enzyme (ACE; rs4343), PAI-1 (rs1799768), brinogen
(rs2070024), ApoA-V (APOA-V; rs662799), cholesterol ester
transfer protein (CETP; rs2303790), fatty acid-binding protein 2
(FABP-2; rs1799883), CRABP-2 (rs2236795), FABP-4 (rs8192688),
ABCG5 (rs6720173), ABCG8 (rs4148211), Apo A-1 (APOA-1; rs670)
and PPAR (rs1800206) SNPs [13] (http://www.genome.gov). PCR
amplied DNA were analysed using enzyme restriction by Iplex
Gold Sequenom technology (coordinacion.cegen@upf.edu).

2.5. Statistical analyses

The sample size calculation was based on the assumption that a
reduction in LDL-C of 0.19 mmol/L [7.35 mg/dL] would be obtained;
equivalent to a reduction of 5% in the LDL-C levels in participants
with LDL-C baseline value of 3.88 mmol/L [150 mg/dL]. With 112
valid participants per group, the study had an 80% statistical power
to detect 0.19 mmol/L [7.35 mg/dL] difference with a standard deviation of 0.51 mmol/L [19.73 mg/dL] and a two-sided type I error
protection level of 5% between the Po-husk and placebo groups.

633

Placebo
Po-husk
Table 2
Lipid prole variables.

0.0004

0.2134

0.0919

<0.0001

0.295 [0.133;0.458] (21.68)

0.024 [0.014;0.061] (3.39)

0.056 [0.009;0.121] (3.15)

0.081 [0.048;0.113] (6.77)

<0.0001
0.46 [0.284;0.637] (6.89)

0.2893
0.034 [0.029;0.096] (2.82)

0.004 [0.039;0.047] (+0.33)

0.03 [0.075;0.015] (2.49)
0.028 [0.094;0.15] (+0.43)
0.433 [0.559;0.306] (6.46)
0.069 [0.044;0.181] (+5.00)
0.227 [0.343;0.11] (16.68)
0.013 [0.013;0.04] (+1.94)
0.01 [0.037;0.017] (1.45)
0.011 [0.056;0.035] (0.62)
0.067 [0.113;0.02] (3.77)
0.006 [0.016;0.029] (+0.52)
0.074 [0.098;0.051] (6.25)
0.325
0.368
0.785
0.712
0.742
0.916
0.138
0.157
0.284
0.303
0.16
0.132

1.201
1.207
6.576
6.699
1.38
1.361
0.67
0.688
1.776
1.779
1.164
1.184

As depicted in Fig. 1, of the participants considered potentially eligible, 209 (82.3%) in the ITT analysis, 187 (73.6%) in the
PP were analysed. Table 1 summarizes the baseline characteristics of participants. In Table 2, the ITT analysis at 8 weeks showed
that the mean LDL-C concentration decreased by 6.1% (mean:
0.29 mmol/L; 95%CI: 0.40 to 0.19) [11.22 mg/dL (15.48 to 7.35)]
in the Po-husk group (P = 0.0002) while, in the placebo group, the
reduction was miniscule (0.1%). At 8 weeks, the rate of subjects
achieving the 5% target reduction in LDL-C was 24.6% higher in the
Po-husk (51.5%; 52 of 101) compared to placebo (26.9%; 29 of 108)
group (P = 0.0004).
In ITT analysis at 8 weeks (Table 2), there were signicant differences in the Po-husk group relative to baseline with respect to
ApoB-100 (P < 0.0001), TC (P < 0.0001), TG (P = 0.0004) and oxLDL
(P = 0.0003).
At 8 weeks (Table 3), no signicant differences between groups
were observed with respect to hs-CRP, IL-6, leucocytes, ICAM-1,
VCAM-1, folic acid and PAI-1 concentrations. However, systolic BP
was lower in Po-husk group compared to placebo (4.013 versus 3.78 mm Hg; 95%CI: 1.2626.765; P = 0.0045). No signicant
differences were found with respect to diastolic BP.
In ITT analysis at 8 weeks, Po-husk reduced insulin plasma
concentration by 7.10 (95%CI: 9.97 to 4.23 pmol/L; P = 0.021)
and HOMA-index by 0.31 (95%CI: 0.42 to 0.20 U/mL*mmol/L;
P = 0.005), without signicant changes in plasma glucose concentrations.
No associations were observed between plasma LDL-C concentrations and any of the DNA polymorphisms studied.
In Po-husk group, plasma TG concentrations were more reduced
in homozygote G and DEL individuals by 3050% (mean SD;

Baseline
Mean SD

3. Results

Rest of lipid prole, lipoproteins and apolipoproteins

Placebo
HDL-C (mmol/L)
Po-husk
Placebo
Total cholesterol
Po-husk
(mmol/L)
Placebo
Triglycerides
Po-husk
(mmol/L)
Placebo
ApoB/ApoA ratio
Po-husk
Placebo
Apolipoprotein A-1
Po-husk
(g/L)
Placebo
Apolipoprotein
Po-husk
B-100 (g/L)

Change at 8 weeks relative to baseline

Adjusted mean [95%CI] (% change from baseline)a

Although 270 participants were sufcient to allow for a 15%

drop-out rate, the study planned to include 290 participants in
order to guarantee participants progressing on to the extension
period with simvastatin.
Results are expressed as mean standard deviation (SD), baseline adjusted least square mean, 95% condence intervals (95%CI)
and percentages.
The continuous efcacy variables were analysed by ANCOVA
(ANalysis of COVAriance) with treatment and baseline values
included in the model. Non-Gaussian continuous variables were
analysed by non-parametric ANCOVA. Categorical variables were
analysed using the Fisher exact test. Primary efcacy analysis was
an intention-to-treat (ITT). Per protocol (PP) and safety analysis
were also performed.
Data from participants who received simvastatin added to Pohusk or to placebo arm from week-8 to week-16 were analysed
(for exploratory purposes only, as planned in the protocol) using
intra-group comparisons.
Data were analysed using the SAS software package version 9.1.3
(SAS Institute Inc., Cary, NC, USA).

0.291 [0.139;0.443] (5.97)

Results are expressed as means standard deviation and frequencies (%). Abbreviations: BMI: body mass index, calculated as weight in kilograms divided by height in
meters squared.

0.008 [0.113;0.098] (0.17)

0.299 [0.408;0.19] (6.14)

54.24 9.96
72.18 13.14
1.64 0.11
26.82 3.72
45 (44.6)
23 (22.8)

4.748 0.641
4.873 0.624

55.59 11.5
72.72 11.24
1.63 0.09
27.45 3.54
46 (42.6)
22 (20.4)

Age (years)
Weight (kg)
Height (m)
Body mass index (kg/m2 )
Gender: male, n (%)
Current smoker, n (%)

Main outcome
LDL-C (mmol/L)

Po-husk (n = 101)

Treatment differences
Adjusted mean [95%CI] (% difference from placebo)a

Placebo (n = 108)

Variable

Variable

0.0002

Table 1
Baseline characteristics of study participants.

Notes: results are expressed as means standard deviation or baseline adjusted least square means [95%CI]. Abbreviations: HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol. To convert SI
units to conventional units (mg/dL): cholesterol multiplies by 38.61; triglycerides multiply by 88.50.
a
Mean relative change = [mean baseline] [adjusted mean at week-8]/[mean baseline].

R. Sol et al. / Atherosclerosis 211 (2010) 630637

634

Table 3
Other than lipid prole efcacy variables.

Blood pressure and bodyweight

Weight (kg)
sBP (mm Hg)
dBP (mm Hg)

Placebo
Po-husk
Placebo
Po-husk
Placebo
Po-husk

Plasma glucose, insulin concentrations and HOMA-index

Placebo
HOMA
Po-husk
(U/mL*mmol/L)
Placebo
Insulin (pmol/L)
Po-husk
Placebo
Glycaemia (mmol/L)
Po-husk

Change at 8 weeks relative to baseline

Adjusted mean [95%CI]

80.395 16.821
80.288 18.869
36.41 27.705
38.026 29.179
11.43 (8.375;17.58)
11.16 (9.27;16.72)
1.885 (0.785;3.54)
1.23 (0.51;2.53)
1.66 (1.25;2.67)
1.74 (1.17;2.38)
6.302 1.565
6.395 1.726
226.35 55.255
229.17 74.16
734.43 181.49
739.31 192.16

1.529 [1.017;4.075]
5.292 [7.925;2.659]
3.571 [7.619;0.477]
6.759 [10.94;2.573]
0.27 (2.14;1.345)
0.87 (2.66;1.06)
0.06 (0.465;0.69)
0.02 (0.33;0.325)
0.03 (0.445;0.6)
0.07 (0.43;0.52)
0.160 [0.09;0.409]
0.04 [0.218;0.298]
1.771 [3.413;6.956]
0.484 [4.877;5.845]
0.709 [19.3;17.887]
7.158 [12.07;26.388]

72.72 11.236
72.18 13.14
126.15 15.422
129.51 17.474
79.63 9.999
81.257 11.084

0.543 [0.863;0.223]
0.839 [1.17;0.508]
0.231 [1.676;2.139]
3.782 [5.755;1.809]
1.442 [2.842;0.041]
2.775 [4.224;1.327]

1.657 0.826
1.656 1.386
50.891 21.737
49.385 30.174
5.182 0.607
5.163 0.686

0.095 [0.201;0.011]
0.311 [0.42;0.202]
2.424 [5.198;0.351]
7.103 [9.972;4.234]
0.015 [0.093;0.063]
0.087 [0.168;0.007]

Treatment differences
Adjusted mean [95%CI]

6.821 [3.158;10.483]

0.0003

3.188 [2.636;9.011]

0.2818
0.1357
0.1412
0.7125

0.119 [0.24;0.478]

0.5129

1.287 [6.172;8.746]

0.7341

7.867 [34.62;18.885]

0.5627

0.295 [0.165;0.756]

0.2075

4.013 [1.262;6.765]

0.0045

1.334 [0.685;3.352]

0.1941

0.215 [0.063;0.368]

0.0058

4.68 [0.688;8.672]

0.0218

0.072 [0.04;0.184]

0.2046

Notes: all results are expressed as means standard deviation and baseline adjusted least square means [95%CI], except for folic acid, high sensitivity C-reactive protein (hs-CRP) and interleukin-6. For the latter variables the
analyses were non-parametric and the estimators were median (25th percentile75th percentile). Plasminogen-activator-inhibitor-1 (PAI); intercellular adhesion molecule-1 (ICAM), vascular adhesion molecule-1 (VCAM-1);
systolic blood pressure (sBP); diastolic blood pressure (dBP).

R. Sol et al. / Atherosclerosis 211 (2010) 630637

Inammatory and endothelial dysfunction marker

Placebo
Oxidised LDL (U/L)
Po-husk
Placebo
PAI-1 (ng/mL)
Po-husk
Placebo
Folic acid (nmol/L)
Po-husk
Placebo
Hs-CRP (mg/L)
Po-husk
Placebo
Interleukin-6
Po-husk
(pm/mL)
Placebo

Leukocytes (109/L)
Po-husk
Placebo
ICAM (ng/mL)
Po-husk
Placebo
VCAM-1 (ng/mL)
Po-husk

Baseline
Mean SD

R. Sol et al. / Atherosclerosis 211 (2010) 630637

635

Table 4
Nutritional prole of low-saturated fatty acid diets in the intention-to-treat population (n = 209).
Diet composition
Total caloric value
(Kcal/d)
Energy requirements
(Kcal/d)
Carbohydrates
(%energy)
Lipids (%energy)
Proteins (%energy)
Saturated fatty acid
(%energy)
Fibre (g/d)

Baseline
Mean SD
Placebo
Po-husk
Placebo
Po-husk
Placebo
Po-husk
Placebo
Po-husk
Placebo
Po-husk
Placebo
Po-husk
Placebo
Po-husk

2021.6
2007.4
1940.3
1935.4
43.5
41.9
35.5
35.8
17.6
17.6
10.3
10.5
21.7
20.8

533.57
589.9
331.7
323.3
6.9
6.8
5.7
5.5
3.5
2.9
2.8
2.5
0.9
1.0

Change at 8 weeks relative to baseline

Adjusted mean [95%CI]
209.2 [286.1;132.4]
230.8 [310;151.7]
4.2 [10.02;18.60]
11.9 [26.79;2.807]
2.6 [1.27;3.961]
4.2 [2.85;5.622]
2.5 [3.733;1.429]
4.4 [5.592;3.22]
0.1 [0.322;0.654]
0.3 [0.107;0.897]
1.8 [2.201;1.49]
2.0 [2.419;1.679]
1.61 [0.091;3.31]
2.18 [0.459;3.898]

Treatment difference
Adjusted mean [95%CI]

21.5 [88.76;131.92]

0.7002

16.2 [4.3;36.877]

0.1204

1.6 [3.558;0.317]

0.1007

1.8 [0.171;3.478]

0.0307

0.2 [0.929;0.47]

0.5187

0.2 [0.31;0.717]

0.4361

0.57[2.999;1.86]

0.6429

Note: all results are expressed as means standard deviation; baseline adjusted least square means [95%CI].

0.44 0.20 and 0.31 0.12 mmol/L, respectively; [38.72 17.60

and 27.28 10.56 mg/dL, respectively] (P = 0.005)) of the PAI-1
gene variants and in homozygote A and G by 2050% (mean SD;
0.85 0.19 and 0.21 0.10 mmol/L, respectively; [74.80 16.72
and 18.48 8.80 mg/dL, respectively] (P = 0.031)) of the FABP-2
gene variants.
At week-8, although the same diet had been prescribed for all
participants, the Po-husk group was observed to have consumed a
lower percentage of lipids compared with placebo (1.8%; 95%CI:
0.1 to 3.5; P = 0.03) (Table 4).
In the ITT analysis at 16 weeks, the number of participants
in the Po-husk was similar to that in placebo [61 of 101 (60%)
versus 70 of 108 (65%), respectively; P = 0.5678]. Simvastatin treatment improved the LDL-C, TG, TC, ApoB-100 concentrations and
the ApoB/ApoA ratio in the treatment as well as the placebo arm
(Fig. 2).
Adherence to Po-husk, placebo and simvastatin treatments were
high, >95%. No differences were found with respect to adverse
effects between treatment groups. Specically, gastrointestinal disorders were 13.6% (n = 16) in placebo and 13.2% (n = 15) in Po-husk,
and only one individual had a relatively serious event (a diagnosis
of type 2 diabetes in the placebo group) but which was considered
not to be related to the study.
Both groups maintained bodyweight during the study.

4. Discussion
Po-husk, the proprietary formulation used in this study and
administered at 14 g/d for 8 weeks to individuals with mildmoderate hypercholesterolemia, induced a mean reduction of 6.1%
in LDL-C plasma concentration. Moreover, 24.6% more Po-husk consumers achieved the target 5% reduction of LDL-C compared to
placebo. The Po-husk LDL-C reduction observed was similar to the
5.56% reduction observed with other soluble bres [35] when
an appropriate daily dose is administrated, especially in individuals in whom LDL-C concentrations are elevated [5]. Also, Po-husk
lowered plasma TC by 6.4% and ApoB-100 by 6.2%.
A considerable lowering (16%) of plasma TG was observed and,
compared to placebo, this was of the order of 21%. The magnitude
of the TG-lowering effect is clinically signicant since the reduction
achieved was similar to that which can be obtained by 1200 mg/d
nicotinic acid (of between 17% and 26%); an agent that has several side effects [13]. In an earlier study with Po-husk administered
at 10.5 g/d, plasma TG was reduced by about 6.7% in men with
ischemic heart disease [6] and, in the present study, the Po-husk
dose is 14 g/d.

In the Po-husk group in the present study, the TG-lowering

effect was magnied, in that the response was about 3050% in
individuals homozygous for G and DEL variants of PAI-1 gene and
2050% in homozygote A and G variants of the FABP-2 gene. If the
high frequency of these gene variants of about 0.5 is conrmed in
a general population, Po-husk consumption could be expected to
induce an optimised plasma TG-lowering response. Thus, Po-husk
has a hypotriglyceridaemic effect compared with other soluble
bres [35] but, surprisingly, the TG reduction was not associated
with increased HDL-C concentrations. The hypotriglyceridaemic
effects of Po-husk diet have been explored in obese Zucker rats.
The proposed mechanism is that of decreased triglyceride and
cholesterol production and accumulation in the liver, resulting
from enhanced AMP-activated protein kinase (AMPK) activation
and acetyl-coA carboxylase and fatty acid synthase inhibition in
this organ. The result is a reduction in circulating free fatty acid
and insulin concentrations [7]. A possible delay in the absorption
of triglycerides and sugars from the small intestine produced by
soluble bre can also be considered an explanatory mechanism
[3].
Additionally, Po-husk lowered the concentration of oxLDL in
plasma; a commonly used marker of oxidative damage involved
in CVD [14,15].
In our study, Po-husk treatment reduced systolic BP by
3.7 mm Hg, which was greater than the 1.3 mm Hg in systolic
BP described for soluble bre in a meta-analysis of randomised
controlled trials [3,16]. Potential mechanisms implicated in the Pohusk lowering of BP could be the attenuation of insulin response
which improves vascular endothelial function [3]. Also, the reduction of oxLDL observed could facilitate the nitric oxide availability
leading to decreased BP [1719]. The magnitude of the hypotensive effect of Po-husk is clinically signicant; the BP reduction of
3.7 mm Hg being close to that usually achieved with pharmacotherapy using -blockers (around 5 mm Hg) [20], but without
the side effects [20].
Po-husk reduced plasma TG, insulin concentrations, systolic
BP, and insulin-sensitivity as measured by HOMA-index [21], and
oxLDL [22]. These changes imply a potential benecial role in
metabolic syndrome clinical status.
The design of the present study was the sequential management (as in the style of primary prevention), of CVD over 6 weeks
and, in which, if the LDL-C remained 3.35 mmol/L [130 mg/dL], a
cholesterol-lowering drug would be prescribed [1,2]. At 16 weeks,
the intra-group combined action of Po-husk and 20 mg/d simvastatin showed similar LDL-C reduction to that of the placebo
plus 20 mg/d simvastatin. However, in the study by Moreyra et
al. [23], when 15 g/d of soluble bre and 10 mg/d of simvastatin

636

R. Sol et al. / Atherosclerosis 211 (2010) 630637

Fig. 2. ITT analysis performed with the week-16 subset of 131 non-responder patients i.e. plasma LDL-C levels 3.35 mmol/L [130 mg/dL] at week-8; 70 of 108 (65%) in
placebo and 61 of 101 (60%) in the Po-husk group. Reduction in LDL-C, ApoB-100, ApoB/ApoA ratio, total cholesterol and triglycerides were statistically signicant. PP analysis
conrmed ITT results. Abbreviations: HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; Apo, apolipoprotein.

were administered simultaneously for 8 weeks, the LDL-C concentration reduction achieved was similar to that achieved by 20 mg/d
simvastatin alone; the type of diet applied had not been specied [23]. A possible explanation for the different results obtained
could be that the low dose simvastatin (10 mg/d) can produce an
additive effect with soluble bre while 20 mg/d simvastatin, being

much more effective in reducing LDL-C concentration, masks the

cholesterol-lowering effect of the bre.
In our study, despite the same dietary intake being prescribed
to all participants, the Po-husk group had, at 8 weeks, consumed
about 2% less lipids; albeit with the same amount of SFA consumption in both groups. This aspect implies a Po-husk palatability effect

R. Sol et al. / Atherosclerosis 211 (2010) 630637

and would suggest that bodyweight changes need to be taken into

account in future studies. This was not possible in our study because
the intervention protocol stipulated that bodyweight of the participants was to be maintained constant.
The level of signicance of the data was conrmed by the PP
population assessment. Po-husk intake was safe and the prevalence
of adverse events was similar to placebo.
The study limitations are that Po-husk intake was over a short
period and did not focus on clinical outcomes. The appropriate daily
dose of soluble bre remains to be dened. Information on adherence to diet in our free-living individual participants was limited.
In conclusion, apart from the target lowering of LDL-C, the
clinically relevant ndings associated with Po-husk treatment
are a combination of effects on plasma TG, TG related to certain gene variants, TC, ApoB-100, oxLDL, insulin and systolic
BP in individuals with mild-moderate hypercholesterolaemia.
Thus, the candidate patients to receive Po-husk would be those
who present a cluster of various CVD risk factors, such as
metabolic syndrome, especially if they have drug intolerances as
well.
Conict of interest
RS received a research grant support from Rottapharm/Madaus
S.L. (Spain) provided directly to the Hospital Universitari Sant Joan
(Reus, Spain). RS had full access to all data and takes full responsibility for the integrity of the data and for the nal manuscript.
AA is an employee of Rottapharm/Madaus, S.L. (Spain). FT and
GD are members of Laboratory of Biostatistics & Epidemiology
(Barcelona, Spain) and their laboratory received nancial support
from Rottapharm/Madaus, S.L. (Spain) for the conduct of the statistical analyses. The funds for the logistics of the study were provided
directly to the participating Hospitals or Primary Care Centres
where the participants were recruited. RS, EB, MC-C, SN, XL, XF,
J-VV, J-M M, M-CA, R-MV, MH have not received honoraria, and all
afrm that they have no conict of interest.
Acknowledgments
We thank the following for their enthusiastic support in the conduct of the study: B. Hansel, L. Charbonnier, M. Divin, H. Tuijtel, C.
Puig, S. Esteve, A. Vzquez, A. Barts, M. Dachs, I. Perez, A. Rubio,
L. Guillen, I. Cardona, P. Lpez, S. Fernandez, L. Pons, C. GonzlezGmez, R. Albaladejo, L. Iniesta, M. Timn, J. Salas, J-C. Vallv, J.
Ribalta, L. Masana.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.atherosclerosis.2010.03.010.

637

References
[1] Graham I, Atar D, Borch-Johnsen K, et al. European guidelines on cardiovascular disease prevention in clinical practice: executive summary. Eur Heart J
2007;28:2375414.
[2] Grundy SM. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III Guidelines. JACC
2004;44:72032.
[3] Anderson JW, Baird P, Davis Jr RH, et al. Health benets of dietary ber. Nutr
Rev 2009;67:188205.
[4] Sierra M, Garca JJ, Fernndez N, Diez MJ, Calle AP. Farmabra group. Therapeutic effects of psyllium in type 2 diabetic patients. Eur J Clin Nutr
2002;56:83042.
[5] Brown L, Rosner B, Willett WW, Sacks FM. Cholesterol-lowering effects of
dietary ber: a meta-analysis. Am J Clin Nutr 1999;69:3042.
[6] Sol R, Gods G, Ribalta J, et al. Effects of soluble ber (Plantago ovata husk)
on plasma lipids, lipoproteins, and apolipoproteins in men with ischemic heart
disease. Am J Clin Nutr 2007;85:115763.
[7] Galisteo M, Morn R, Rivera L, et al. Plantago ovata husks-supplemented diet
ameliorates metabolic alterations in obese Zucker rats through activation of
AMP-activated protein kinase. Comparative study with other dietary bers.
Clin Nutr 2009;(September) [Epub ahead of print].
[8] Anderson JW, Randles KM, Kendall CWC, Jenkins DJA. Carbohydrate and ber
recommendations for individuals with diabetes: a quantitative assessment and
meta-analysis of the evidence. J Am Coll Nutr 2004;23:517.
[9] Salas-Salvad J, Farrs X, Luque X, et al. Effect of two doses of a mixture of
soluble bres on body weight and metabolic variables in overweight or obese
patients: a randomised trial. Br J Nutr 2008;99:13807.
[10] Ordovas JM, Tai ES. Why study geneenvironment interactions? Curr Opin
Lipidol 2008;19:15867.
[11] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of
low-density lipoprotein cholesterol in plasma, without use of the preparative
ultracentrifuge. Clin Chem 1972;18:499502.
[12] Matthews DR, Hosker JP, Rudenski AS, et al. Homeostasis model assessment:
insulin resistance and beta-cell function from fasting plasma glucose and
insulin concentrations in man. Diabetologia 1985;28:4129.
[13] Brunzell JD, Hypertriglyceridemia. N Engl J Med 2007;357(10):100917.
[14] Meisinger C, Baumert J, Khuseyinova N, Loewel H, Koenig W. Plasma oxidized
low-density lipoprotein, a strong predictor for acute coronary heart disease
events on apparently healthy, middle-aged men from the general population.
Circulation 2005;112:6517.
[15] Fit M, Guxens M, Corella D, et al. Effect of a traditional Mediterranean diet
on lipoprotein oxidation: a randomized controlled trial. Arch Intern Med
2007;167:1195203.
[16] Whelton SP, Hyre AD, Pedersen B, et al. Effect of dietary ber intake on blood
pressure: a meta-analysis of randomized, controlled clinical trials. J Hypertens
2005;23:47581.
[17] Landmesser U, Drexler H. Endothelial function and hypertension. Curr Opin
Cardiol 2007;22:31620.
[18] Guxens M, Fit M, Martnez-Gonzlez MA, et al. Hypertensive status and
lipoprotein oxidation in an elderly population at high cardiovascular risk. Am
J Hypertens 2009;22:6873.
[19] Ceriello A. Possible role of oxidative stress in the pathogenesis of hypertension.
Diabetes Care 2008;31(Suppl. 2):S1814.
[20] Morgan TO, Anderson AI, MacInnis RJ. ACE inhibitors, beta-blockers, calcium
blockers, and diuretics for the control of systolic hypertension. Am J Hypertens
2001;14:2417.
[21] Tuan C-Y, Abbasi F, Lamendola C, McLaughlin T, Reaven G. Usefulness of plasma
glucose and insulin concentrations in identifying patients with insulin resistance. Am J Cardiol 2003;92:60610.
[22] Holvoet P, Lee DH, Steffes M, Gross M, Jacobs Jr DR. Association between
circulating oxidized low-density lipoprotein and incidence of the metabolic
syndrome. JAMA 2008;299:228793.
[23] Moreyra AE, Wilson AC, Koraym A. Effect of combining psyllium ber with
simvastatin in lowering cholesterol. Arch Intern Med 2005;165:11616.