PHYTOTHERAPY RESEARCH

Phytother. Res. 28: 33–41 (2014)

Published online 25 February 2013 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/ptr.4950

Metabolic Management in Overweight Subjects

with Naive Impaired Fasting Glycaemia by
Means of a Highly Standardized Extract From
Cynara scolymus: A Double-blind, Placebo-
controlled, Randomized Clinical Trial

Mariangela Rondanelli,1 Annalisa Opizzi,1 Milena Faliva,1 Patrizio Sala,1 Simone Perna,1
Antonella Riva,2 Paolo Morazzoni,2 Ezio Bombardelli2 and Attilio Giacosa3
1
University of Pavia, Health Sciences Department, Section of Human Nutrition, Azienda di Servizi alla Persona, Via Emilia n.12, Pavia,
Italy
2
Indena S.p.A., Milan, Viale Ortles n. 12, Italy
3
Department of Gastroenterology, Policlinico di Monza, Via Amati 111, Milan, Italy

The aim of this study is to evaluate the efficacy of a dietary supplementation with an extract from Cynara
scolymus (Cs) on the glucose pattern in a group of patients with naïve impaired fasting glycaemia (IFG). A ran-
domized, double-blind, placebo-controlled trial has been performed in 55 overweight subjects with IFG (fasting
blood glucose [FBG]: 6.11  0.56 mmol/l). These subjects were randomly assigned to supplement their diet with
either an extract from Cs (600 mg/d) (26 subjects) or placebo (29 matched subjects) for 8 weeks. The decrease of
FBG was the primary endpoint. The assessment of Homeostatic Metabolic Assessment (HOMA), glycosylated
haemoglobin, A1c-Derived Average Glucose (ADAG), lipidic pattern and anthropometric parameters were
the secondary endpoints. The within groups and percent changes from baseline were analyzed by the signed
rank test. The comparison between groups was performed by Wilcoxon’s two sample test. The supplemented
group had significant decreases of: FBG ( 9.6%), HOMA ( 11.7%), glycosylated haemoglobin ( 2.3%),
ADAG ( 3.1%) and lipidic pattern. The placebo group did not show any significant difference. Compared with
the placebo, the supplemented group showed a significant difference in FBG, HOMA and lipidic pattern. These
data demonstrate the efficacy of Cs extract on the reduction of glycometabolic parameters in overweight
subjects with IFG. Copyright © 2013 John Wiley & Sons, Ltd.

Keywords: dietary supplement; overweight; impaired fasting glycaemia; Cynara scolymus.

et al., 2001). Since long time, traditional phytotherapy

INTRODUCTION
Impaired fasting glycaemia (IFG) is a metabolic condition
which increases the risk of developing type 2 diabetes
(Unwin et al., 2002), one of major burdens in health care
system in many industrialized countries (Shaw et al.,
2010). A survey carried out in Copenhagen showed that
24.7% of males and 17% of females aged 60 years or more
had IFG, indicating an increase of type 2 diabetes of 70%
and 40%, respectively, over a 10-year period, as com-
pared to normoglycaemic subjects (Glümer et al., 2003).
The proper management of IFG is the most cost-
effective way to avoid the many sequelae of diabetes
and primarily cardiovascular diseases (CVDs). If left un-
treated, most of the individuals with IFG will develop
type 2 diabetes within 10 years, and a high proportion of
them will suffer from micro- or macrovascular pathology
(DECODE Study Group, the European Diabetes Epi-
demiology Group, 2001; Eastman et al., 1997; Saydah
Correspondence to: Mariangela Rondanelli, Department of Applied
Health Sciences, Section of Human Nutrition and Dietetics, Faculty of
Medicine, University of Pavia, Endocrinology and Nutrition Unit, Azienda
di Servizi alla Persona di Pavia, University of Pavia, Via Emilia 12, 27100
Pavia, Italy. E-mail: mariangela.rondanelli@unipv.it
Copyright © 2013 John Wiley & Sons, Ltd.
has been used to treat IFG, as an alternative remedy to
glucose-lowering drugs due to their frequent side effects,
such as hypoglycaemia, lactic acidosis, idiosyncratic liver
cell injury, permanent neurological deficit, digestive dis-
comfort, headache and dizziness (Neustadt and Pieczenik,
2008). Hypoglycaemic herbal extracts are widely used as
non-prescription treatment for diabetes (Hui et al.,
2009). In particular, recent in vitro and animal studies
support the hypoglycaemic effect of Cynara scolymus
(Cs) extracts (Arion et al., 1997; Matsui et al., 2006;
Fantini et al., 2010). This activity seems to be mainly
related to the content of chlorogenic acid in Cs (Arion
et al., 1997; Matsui et al., 2006; Fantini et al., 2010;
Karthikesan et al., 2010). This compound is a potent
inhibitor of glucose 6-phosphate translocase, an essential
component of the hepatic glucose 6-phosphatase system
which regulates the homeostasis of blood glucose (Arion
et al., 1997). In addition, dicaffeoylchinic acid derivatives
of Cs can also play a hypoglycaemic role in modulating
the activity of alpha-glucosidase and consequently the
catabolism of dietary carbohydrates (Matsui et al., 2006).
The favourable effects of Cs extracts on serum
glucose regulation have been clearly demonstrated in
animal studies, but intervention studies in humans
are limited.
Received 28 March 2012
Revised 18 December 2012
Accepted 14 January 2013
34 M. RONDANELLI ET AL.
Only very recently, a highly standardized extract
obtained from Cs flowering buds has been investigated
(in combination with an extract from Phaseolus
Vulgaris) in overweight subjects: after 8 weeks of treat-
ment, a higher and significant reduction in glucose net
change was observed in the supplemented group as
compared to controls (Rondanelli et al., 2011).
Given this background, we investigated whether the
highly standardized Cs extract could improve glycaemic
control and insulin sensitivity in overweight subjects
with newly diagnosed IFG.
MATERIALS AND METHODS
Study design. A randomized, double-blind, placebo-
controlledtrial was conducted in overweight and obese
(body mass index (BMI) 25–35 kg/m2) men and women
with newly detected IFG, as defined by American Diabetes
Association (American Diabetes Association and
National Institute of Diabetes, Digestive and Kidney
Diseases, 2002).
The subjects were recruited from Pavia municipality by
advertisement on a local newspaper and were screened by
means of a clinical evaluation and plasma glycometabolic
assessment. The study was carried out at the Dietetic and
Metabolic Unit of the “Santa Margherita’ Institute,
University of Pavia, Italy.
Fifty-five subjects, aged 18–60 years with BMI ranging
from 25 to 35 kg/m2, with IFG (6.1–7.0 mmol/L), glyco-
sylated haemoglobin <7.0% and no history of CVD,
volunteered for the trial (Fig. 1). The subjects were not
taking any medication likely to affect glucose or lipid
metabolism (oral hypoglycaemic agents and statins)
and were free of overt liver, renal and thyroid disease.
The subjects who smoked or drank more than two
standard alcoholic beverages/day (20 g of alcohol/day)
were excluded from the study. Physical activity was
recorded. Sedentary subjects were admitted to the
study. The experimental protocol was approved by the
Ethics Committee of the University of Pavia, and all
the volunteers gave their written informed consent.
Figure 1. Flow diagram of a trial of supplementation versus pla-
cebo in the treatment of healthy overweight subjects.
Copyright © 2013 John Wiley & Sons, Ltd.
Twenty-six subjects were randomized to Cs supplement
and 29 to placebo. As subjects were enrolled, they were
assigned a progressive subject number.
Identical products for each treatment group were
assigned to a subject number according to a coded
(AB) block randomization table prepared by an
independent statistician. Investigators were blinded to
the randomization table, the code assignments and the
procedure. Independently by treatment (Cs supplement
or placebo), the subjects followed a similar low-energy
diet. The dietary treatment was associated to three daily
oral assumptions (before breakfast, lunch and dinner) of
film-coated tablets of 200 mg of standardized Cs flower-
ing buds extract or placebo. The tablets were manufac-
tured by Indena, Milan - Italy. The supplementation
period was 8 weeks.
The Cs flowering heads extract was prepared through
alcoholic extraction (EtOH 70%), and the extract was
characterized by a high content in caffeoylquinic acid
(high-performance liquid chromatography [HPLC] - %
w/w between 30 and 60%) and flavonoids (expressed as
luteolin glycosides; HPLC - % w/w between 2 and 5%).
The yield expressed as drug/extract ratio is 120/1. Details
of preparation have been described elsewhere (Fantini
et al., 2010).
Compliance to the supplementation regimen was
defined as the number of tablets actually taken by each
subject, divided by the number of tablets that should have
been taken over the course of the study. Adverse events
(AEs) were based on spontaneous reporting by subjects
as well as open-ended inquiries by members of the re-
search staff. There were no study dropouts (Fig. 1).
Glycaemic and lipidic parameters. The glycaemic and
lipidic parameters were assessed before the start of the
study and at the end of treatment (EoT). Moreover,
the fasting blood glucose (FBG) was evaluated 60 days
after the end of the intervention. In order to avoid
venipuncture stress, blood samples were obtained
through an indwelling catheter inserted in an antecubi-
tal vein. Blood samples were immediately centrifuged
and stored at 80  C until assayed. FBG, total choles-
terol (TC), low-density lipoprotein-cholesterol (LDL-
C), high-density lipoprotein-cholesterol (HDL-C) and
triglyceride (TG) levels were measured by automatic
biochemical analyzer (Hitachi 747, Tokyo, Japan).
Serum concentration of haemoglobin A1c (HbA1c)
was determined by high-performance liquid chromato-
graphic method using automatic HbA1c analyzer
(Tosoh HLC-723G7, Japan). A1c-Derived Average
Glucose (ADAG) was calculated (Nathan et al., 2008).
The serum insulin was evaluated by a double antibody
RIA (Kabi Pharmacia Diagnostics AB, Uppsala,
Sweden) and expressed as pmol/L. The intra- and
inter-assay coefficients of variation were below 6%,
and the low detection limit was 10.7 pmol/L. To determine
insulin resistance, subjects were instructed to fast for 12 h
before obtaining the blood sample. Furthermore, the
subjects refrained from any form of physical exercise
for 48 h before the study. Female subjects were tested
during the early follicular phase of their menstrual
cycles (days 3–10). Insulin resistance was evaluated
using the Homeostasis Model Assessment (HOMA)
(Haffner et al., 1996). Finally, for the assessment of
safety, routine blood biochemistry parameters (blood
count, serum protein elecrophoresis, creatinine, liver
Phytother. Res. 28: 33–41 (2014)
 CYNARA SCOLYMUS AND IMPAIRED GLYCAEMIA
and thyroid function) were evaluated at the start and at
the end of intervention.
Anthropometric measurements and dietary counselling.
Nutritional status was assessed using anthropometric
measurements before the start of the study and at EoT.
Body weight and height were measured, and the BMI
was calculated (kg/m2). Skinfold thicknesses (biceps,
triceps, suprailiac, subscapular) were measured twice
using a harpender skinfold caliper at 5 min intervals in
each site following a standardized technique (Frisancho,
1984). Sagittal abdominal diameter at the L4–5 level in
the supine position and waist girth was also measured.
Anthropometric parameters were always collected by
the same investigator.
Subjects were trained to follow a regimen that main-
tained a prudent balance of macronutrients: 28% of en-
ergy from fat (cholesterol < 200 mg), 57% of energy
from carbohydrates (10% from simple carbohydrates),
with 20–25 g of bran, and 15% of energy from protein.
A registered dietician performed initial dietary counsel-
ling. Subjects were trained to restrict their daily energy
intake by a moderate amount, 3344 kJ/d less than daily
requirements based on WHO criteria (World Health
Organization, 1985), with a regimen that maintained a
prudent balance of macronutrients: 25–30% of energy
from fat (cholesterol < 200 mg), 55–60% of energy
from carbohydrates (10% from simple carbohydrates),
with 25 g of bran and 15–20% of energy from protein.
A registered dietician performed initial dietary coun-
selling. A 3-day weighed-food record of 2 weekdays
and 1 weekend day was performed during the first and
the last weeks of the study. Dietary records were ana-
lyzed using a food-nutrient database (Rational Diet,
Milan, Italy).
Satiating capacity. The satiating capacity was assessed
using Haber’s scale which is an analogue visual scale
ranging from –10 (representing extreme hunger: pain-
fully hungry) to +10 (representing extreme satiety:
full to nausea) (Haber et al., 1977). Precisely, the
subjects indicated the level of agreement in respect
to hunger or satiety by pointing to an appropriate
place along the graduated visual scale. The test was
performed every day before lunch time by all the
subjects included in the study during the entire period
of treatment.
Psychodynamic pattern. A Beck Depression Inventory
(BDI-II) was taken to assess depressive symptoms: a
score of 10 to 30 was indicative of depressive symptoms
(Steer et al., 1999). The BDI-II was performed by all the
subjects at the start of the study and at EoT. The psycho-
dynamic test was conducted under standardized condi-
tions of comfort and silence, with a study technician
always in attendance.
Statistical methods. All statistical analyses were per-
formed by the SAS System version 9.2, choosing the 5%
(p ≤ 0.05) as threshold value of statistical significance.
The primary endpoint of the study was the supplemented
versus placebo comparison of the change in fasting glucose
blood levels between baseline and EoT. This was a single
comparison; thus, there were no multiplicity concerns. In
the case of the secondary endpoints, no adjustment of the
5% significance level was done for multiple comparisons/
Copyright © 2013 John Wiley & Sons, Ltd.
35
multiplicity due to their descriptive nature. Secondary effi-
cacy endpoints were the changes between baseline and
EoT of the other parameters of glycidic metabolism,
lipidic pattern, psychodynamic test and anthropometric
parameters.
Fasting glycaemia was also evaluated at the end of the
follow-up (EoF) that is 60 days after the EoT. Differences
in gender distribution between the two groups were
assessed by the chi-square test. Due to the small sample
size, distribution-free non-parametric tests have been
adopted to analyze the discrete or continuous data. The
within groups changes and percent changes from baseline
were analyzed by the signed rank test (SR). At each as-
sessment time, the comparison between groups was per-
formed by the Wilcoxon’s (W) two-sample test. The
Haber scores recorded daily were averaged every ten
days, producing six assessment occasions: days 1–10;
11–20; 21–30; 31–40; 41–50; 51–60. The average of the
scores recorded over the two pre-treatment weeks was
used as baseline. The pattern over time of the scores
of each group was assessed by the Friedman’s test.
The comparisons within groups versus baseline and be-
tween groups at each assessment occasion were per-
formed by the SR and W tests, respectively. It was
also performed analysis of covariance with the delta fast-
ing glycaemia, as an independent variable, and the treat-
ment and delta BMI, as explanatory variables. The same
model of ANCOVA was also applied to the delta of
every lipidic parameter.
Safety was assessed by laboratory tests performed at
baseline and EoT and by recording volunteered AEs.
RESULTS
All the 55 subjects completed the 60 days intervention
trial and their characteristics at the baseline are shown
in Table 1; the placebo and supplemented groups were
homogeneous for all glyco-lipidic parameters.
Glucose metabolism
The decrease in FBG was statistically significant in the
supplemented group ( 9.6%, p < 0.001), whereas it
was not statistically significant in the placebo group. At
the end of intervention, the difference between groups
was statistically significantly (p = 0.001). The post-hoc
power based on the observed treatments difference at
EoT, with a pooled SD of 0.50 mmol/L and Alpha = 0.05,
reaches the 99%. The changes from baseline in the insu-
lin levels of two groups did not significantly differ. A
statistically significant between groups difference was
found in the percent changes from baseline (p = 0.044).
At the end of intervention, the supplemented group
showed a significant decrease in the glycosylated
haemoglobin ( 2.32%, p = 0.003), a significant decrease
in the ADAG ( 3.6%, p = 0.002) and a significant de-
crease in HOMA index (p < 0.001), whereas the placebo
group did not show significant changes. These results
have been shown in Table 2. There was no significant
correlation within each group between delta BMI and
fasting glycaemia: supplemented group (Pearson
r = 0.074, p = 0.721; Spearman r = 0.003, p = 0.989); pla-
cebo group (Pearson r = 0.028, p = 0.888; Spearman
Phytother. Res. 28: 33–41 (2014)
36 M. RONDANELLI ET AL.

Table 1. Demographics and baseline characteristics of the studied patients

Supplemented Placebo Total

Variablea (N = 26) (N = 29) (N = 55) Pb

Gender F 15 (58%) 15 (52%) 30 (54.5%) 0.657

M 11 (42%) 14 (48%) 25 (45.5%) (w2)
Age Years 54.7 [11.6] 53.6 [8.2] 54.1 [9.8] 0.821
BMI kg/m^2 31.74 [4.30] 29.77 [3.08] 30.70 [3.80] 0.079
Glucose mmol/L 6.049 [0.498] 6.172 [0.607] 6.114 [0.556] 0.879
Insulin mcU/ml 13.50 [5.33] 12.83 [6.71] 13.14 [6.05] 0.241
HOMA index 3.671 [1.702] 3.542 [1.799] 3.603 [1.739] 0.474
Glycosylated haemoglobin (%) 6.59 [0.43] 6.44 [0.74] 6.50 [0.63] 0.285
(%) Haemoglobin
ADAG 142.4 [12.4] 138.3 [21.2] 140.0 [18.1] 0.285
Total cholesterol (mmol/L) mmol /L 5.903 [1.018] 6.050 [0.794] 5.982 [0.899] 0.415
(mmol/L) (mmol/L)
(mmol/L) (mmol/L)
HDL-cholesterol mmol /L 1.384 [0.430] 1.529 [0.730] 1.462 [0.608] 0.386
(mmol/L)
Total/HDL-cholesterol ratio mmol /L 4.621 [1.636] 5.055 [2.552] 4.854 [2.168] 0.755
LDL-cholesterol mmol /L 3.976 [0.883] 3.886 [0.449] 3.928 [0.680] 0.430
(mmol/L)(mmol/L)
Triglycerides mmol /L 1.395 [0.559] 1.380 [0.356] 1.387 [0.459] 0.686
BDI-II scale score 8.35 [7.68] 7.93 [6.91] 8.13 [7.22] 0.979
Haber scale score 1.630 [2.033] 1.880 [1.550] 1.760 [1.780] 0.826
a
data are expressed as mean values and [SD], if otherwise not specified
b
Wilcoxon’s two-sample test, if otherwise not specified

r = 0.111, p = 0.575), as shown in Fig. 2. Moreover, the
results of ANCOVA analysis demonstrated that changes
in BMI did not affect changes in fasting glycaemia.
Even, the interaction between treatment and BMI was
not significant, indicating that the regression between
delta fasting glycaemia and delta BMI was similar in
both groups, as shown in Table 3.
Lipid parameters
TC significantly decreased in the supplemented group at
the EoT ( 6.30%, p = 0.001), but it did not change in
the placebo group. HDL-C did not show any significant
change from baseline at the EoT either in the supple-
mented group or in the placebo group. The ratio be-
tween total and HDL cholesterol significantly
decreased from baseline in the supplemented group
( 5.69%, p = 0.041). At EoT, blood LDL-C level signifi-
cantly decreased in the supplemented group ( 10.57%,
p = 0.002), while it did not change significantly in the
placebo group. TGs did not significantly change from
baseline in both the intervention and control group.
These results have been shown in Table 4. The
ANCOVA results are in Table 3. They show that there
was a significant relationship between decrease in TC
and decrease in BMI (p = 0.039), but regression between
delta TC and delta BMI was different in the two groups
(interaction treatment by delta BMI, p = 0.049). The by
group correlation between changes in TC and changes
in BMI did not evidence any statistically significant
relationship significant: supplemented group (Pearson
r = 0.308, p = 0.135; Spearman r = 0.267, p = 0.198); pla-
cebo group (Pearson r = 0.086, p = 0.665; Spearman
r = 0.105, p = 0.595). A significant relationship was also
found between changes in LDL-C and changes in BMI

Table 2. BMI and glucose metabolism. Mean changes from baseline

Supplemented mean Placebo mean Treatment effect

change and change and mean difference and
95%CI 95%CI 95%CI
Variable p value
Mean 95%CI Mean 95%CI Mean 95%CI (W)

BMI (kg/m^2) 0.95} 1.27 0.64 0.41 1.45 0.63 0.54 1.62 0.53 0.001
Glucose (mmol/L) 0.62} 0.87 0.36 0.00 0.14 0.13 0.61 0.89 0.33 0.001
Insulin (mcU/ml) 0.47 1.21 0.28 0.45 1.06 1.97 0.92 2.58 0.74 0.057
Glycosylated haemoglobin 0.16# 0.25 0.06 0.08 0.24 0.08 0.08 0.26 0.11 0.153
ADAG 4.45# 7.22 1.68 2.28 6.82 2.27 2.17 7.38 3.04 0.178
HOMA index 0.54} 0.84 0.23 0.15 0.33 0.62 0.68 1.24 0.13 0.006

W = Wilcoxon’s test
*p < 0.05; # p < 0.01; } p < 0.001 at the signed rank test

Copyright © 2013 John Wiley & Sons, Ltd. Phytother. Res. 28: 33–41 (2014)
 CYNARA SCOLYMUS AND IMPAIRED GLYCAEMIA 37

r = 0.167, p = 0.425); placebo group (Pearson r = 0.186,

p = 0.343; Spearman r = 0.043, p = 0.830).

Anthropometric parameters

At EoT, the decrease in BMI was statistically significant in

Figure 2. Scatter diagram of changes in blood glucose versus
changes in BMI.
(p = 0.034), but in this case, the interaction ‘Treatment
by delta BMI’ was not significant (p = 0.065), the by
groups correlation results being the following: supple-
mented group (Pearson r = 0.315, p = 0.125; Spearman
the supplemented group, while it did not significantly
change in the placebo group, the between groups differ-
ences in changes from baseline being statistically significant
(p = 0.001). At EoF, the differences versus baseline were
statistically significant in the two groups, but the changes
from baseline were more pronounced in the supplemented
group when compared with placebo (p = 0.019 and
p = 0.021 for changes and percent changes, respectively).
The results regarding BMI have been reported in Table 2.
As concerns skinfold thickness, statistically significant
reductions of baseline values of triceps, subscapular and
suprailiac zones have been observed in the supplemented
group whereas no statistically significant changes were
found in the placebo group. Between groups statistically
significant differences in the changes from baseline were
observed for the subscapular zone (p = 0.007 for changes
from baseline, p = 0.011 for percent changes from base-
line). As concerns the other anthropometric parameters,
statistically significant reductions of the baseline values
have been observed in the supplemented group for arm

Table 3. Results of ANCOVA

Variable Source of variation DF Sum of squares Mean square F value Pr > F

Treatment 1 2.30078 2.30078 8.73 0.005

Changes in glucose Changes in BMI 1 0.04311 0.04311 0.16 0.688
Treatment* Changes in BMI 1 0.05558 0.05558 0.21 0.648
Treatment 1 0.18584 0.18584 0.45 0.506
Changes in total cholesterol Changes in BMI 1 1.86721 1.86721 4.50 0.039
Treatment* Changes in BMI 1 1.69442 1.69442 4.08 0.049
Treatment 1 0.01858 0.01858 0.59 0.445
Changes in HDL-cholesterol Changes in BMI 1 0.00003 0.00003 0.00 0.976
Treatment* Changes in BMI 1 0.00333 0.00333 0.11 0.746
Changes in the ratio Treatment 1 0.10645 0.10645 0.18 0.670
Total chol./HDL-chol. Changes in BMI 1 1.12014 1.12014 1.93 0.171
Treatment* Changes in BMI 1 0.98676 0.98676 1.70 0.198
Treatment 1 0.57987 0.57987 1.37 0.248
Changes in LDL-cholesterol Changes in BMI 1 2.01717 2.01717 4.75 0.034
Treatment* Changes in BMI 1 1.50812 1.50812 3.55 0.065
Treatment 1 0.00037 0.00037 0.00 0.958
Changes in triglycerides Changes in BMI 1 0.00669 0.00669 0.05 0.822
Treatment* Changes in BMI 1 0.06262 0.06262 0.48 0.493

Table 4. Lipids metabolism. Mean changes from baseline

Supplemented mean Placebo mean Treatment effect mean

change and change and difference and
95%CI 95%CI 95%CI
Variable p value
Mean 95%CI Mean 95%CI Mean 95%CI (W)

Total cholesterol (mmol/L) 0.44# 0.82 0.06 0.07 0.03 0.18 0.52 0.91 0.13 0.001
HDL-cholesterol (mmol/L) 0.00 0.07 0.07 0.04 0.11 0.03 0.04 0.06 0.13 0.399
Col/HDL ratio 0.35* 0.72 0.02 0.01 0.23 0.25 0.37 0.78 0.05 0.096
LDL-cholesterol (mmol/L) 0.50# 0.87 0.13 0.13 0.01 0.27 0.63 1.02 0.24 <0.001
Triglycerides (mmol/L) 0.10 0.30 0.10 0.04 0.10 0.02 0.06 0.27 0.14 0.209

W = Wilcoxon’s test
*p < 0.05; # p < 0.01

Copyright © 2013 John Wiley & Sons, Ltd. Phytother. Res. 28: 33–41 (2014)
38 M. RONDANELLI ET AL.

circumference, AMA, AFA, MAC, waist, hips, calf and in
the placebo group for brachial circumference and waist-hip
ratio (WHR). The reductions versus baseline in the supple-
mented group were significantly more marked than pla-
cebo for: hips (p = 0.005 for both changes and percent
changes from baseline) and calf (p = 0.020 for changes from
baseline, p = 0.015 for percent changes from baseline); on
the other hand, the reduction in WHR observed in the pla-
cebo group was significantly higher than in the supplemen-
ted group (p < 0.001 for both changes and percent changes
from baseline). The results of the statistical analysis applied
to the changes from baseline of anthropometric para-
meters are reported in Table 5.
Satiating capacity, psychodynamic pattern
The Haber test scores of the supplemented group did not
show any statistically significant trend on time (Friedman
test: 5.522; p = 0.479), while the scores of placebo group
showed a statistically significant trend on time (Friedman
test: 24.511; p < 0.001), the changes from baseline being
significant in this group at ‘Days 11–20’ (p = 0.007). No be-
tween groups statistically significant differences in Haber
scores have been detected. The changes from baseline
analyzed at EoT are shown in Table 5.
As regards BDI, the scores of the two groups did not
show any statistically significant change from baseline to
EoT. No statistically significant differences between
groups have been detected (Table 5).
Table 6 showed the results of a 3-day weighed-food
record of 2 weekdays and 1 weekend day performed
during the first week of intervention and during the last
week of intervention in the two groups. No significant
differences between treatment groups were shown.
AEs
No AEs were reported during the study in either group,
and the values in routine blood biochemical parameters
evaluated for the safety of the two groups did not show
any statistically significant changes from baseline, nor
statistically significant differences between groups.
DISCUSSION
The results of this double-blind, placebo-controlled,
randomized clinical trial demonstrate that the treatment
with a highly standardized extract from Cs flowering
heads shows favourable glyco-metabolic effects in over-
weight subjects with naïve IFG. This is proven by the
significant reduction of baseline values of FBG after
60 days of intervention with Cs extract, and this was
the primary efficacy endpoint of the study. Moreover, a
significant reduction of glycosylated haemoglobin,
ADAG and HOMA has also been observed in the inter-
vention group, thus confirming the efficacy of Cs extract
in the control of glucose metabolism in IFG patients. On
the contrary, insulinemia did not change significantly
after treatment with Cs extract, and this is an expected
observation due to the fact that insulin basal levels were
within normal limits in our group of subjects. As a con-
sequence, the significant reduction of HOMA observed
only in the intervention group at the end of the study
appears mainly correlated to the reduction of glycaemia.
HOMA is an indirect index of insulin resistance, and its
reduction supports the efficacy of supplementation with
Cs extract in preventing the development of type 2
diabetes (Unwin et al., 2002).
This effect on glucose reduction is in agreement
with the use of Cs extracts in traditional medicine as
an antidiabetic remedy (Hernandez-Galicia et al.,
2002). As a matter of fact, it is well known that the
great problem of traditional herbal treatments is, in
most of the cases, the lack of well standardized
products and the lack of scientific evidence of efficacy,
obtained by means of controlled clinical trials. On the
contrary, this study has been conducted with a highly
standardized Cs extract and with a double-blind,

Table 5. Anthropometric, satiety and depression parameters. Mean changes from baseline

Supplemented mean Placebo mean Treatment effect mean

change and change and difference and
95%CI 95%CI 95%CI
Variable p value
Mean 95%CI Mean 95%CI Mean 95%CI (W)

Skinfold thickness: Biceps 0.96 1.93 0.0 0.95 2.06 0.17 0.01 1.46 1.43 0.648
Skinfold thickness: Triceps 0.81* 1.49 0.12 0.48 1.26 0.30 0.33 1.35 0.70 0.273
Skinfold thickness: Subscapular 1.40# 2.38 0.42 0.61 0.18 1.39 2.01 3.22 0.79 0.003
Skinfold thickness: Suprailiac 0.65* 1.27 0.04 0.41 1.76 0.94 0.24 1.73 1.24 0.271
Brachial circumference 1.06} 1.48 0.63 0.39* 0.71 0.07 0.66 1.18 0.15 0.002
AMA 3.13# 5.13 1.14 0.76 1.58 0.06 2.37 4.42 0.32 0.025
AFA 0.36} 0.58 0.14 0.05 0.17 0.06 0.31 0.54 0.07 0.006
MAC 0.80} 1.28 0.33 0.24 0.49 0.01 0.56 1.07 0.05 0.045
Waist 1.71} 2.55 0.88 0.71 1.73 0.30 1.00 2.29 0.30 0.102
Hips 1.63} 2.23 1.04 0.16 0.88 0.56 1.47 2.40 0.55 0.001
WHR 0.00 0.01 0.00 0.08} 0.11 0.06 0.08 0.06 0.11 0.000
Calf 0.65} 0.89 0.41 0.13 0.32 0.07 0.53 0.83 0.23 0.000
BDI-II 0.08 1.77 1.92 0.18 0.75 0.39 0.26 1.57 2.08 0.780
HABER score 0.09 0.85 1.03 0.38 1.32 0.56 0.47 0.83 1.77 0.443

W = Wilcoxon’s test
*p < 0.05; # p < 0.01; } p < 0.001 at the signed rank test

Copyright © 2013 John Wiley & Sons, Ltd. Phytother. Res. 28: 33–41 (2014)
 CYNARA SCOLYMUS AND IMPAIRED GLYCAEMIA 39

(23.9); 52.3 (2.5)

(23.9); 34.0 (4.4)
(10.9); 18.3 (3.4)
placebo-controlled, randomized design. The issue of

66.0 (5.9); 18.2 (1.0)

47.6 (5.5); 29.5 (2.1)
12.9 (3.0); 8.0 (1.1)
standardization, characterization, preparation and
Last week of
intervention
toxicity of an herbal extract is crucial (Shan et al.,

(placebo)
2007), and, as far as Cs extracts are considered, a rele-

6083 (567.2)

(33.0)
(5.1)
vant diversity has been shown when different dietary
supplements were compared (Schütz et al., 2006).

202.6
131.6
71.2
21.5
184.9
The Cs extract used in this study was highly standar-
dized and characterized by a high content in caffeoyl-
quinic acid and flavonoids, expressed as luteolin
glycosides (Fantini et al., 2010). The daily doses of
Table 6. A 3-day weighed-food record of 2 weekdays and 1 weekend day was performed during the first week of intervention and during the last week of intervention

its polyphenolic components is in the range of the

average food consumption in the Mediterranean
population and in line with the one used in a recent
study that evaluated the absorption and metabolism
of bioactive molecules after oral consumption of
(24.4); 54.8 (2.0)
(23.3); 39.1 (3.0)
(12.5); 15.7 (2.9)
76.2 (6.6); 16.8 (1.1)
57.0 (5.2); 28.3 (2.0)
13.8 (1.4); 6.9 (0.9)

cooked edible heads of Cs in human subjects (Azzini

*Mean (SD); Nutritional evaluation from Carnovale E, Marletta L: ‘Tabelle di composizione degli alimenti’, Istituto Nazionale della Nutrizione, Roma 1997
et al., 2007). Furthermore, this study showed biologic-
First week of
intervention
(placebo)

ally effective concentrations of artichoke markers in

7580 (559.9)

(32.4)

the human plasma (Azzini et al., 2007).

(4.2)

The mechanism of action of Cs extracts on glycaemia

is multifactorial, as shown by various in vitro and animal
264.6
189.0
75.6
25.7
192.1

studies (Arion et al., 1997; Matsui et al., 2006; Karthikesan

et al., 2010; Hemmerle et al., 1997; Arion et al.,
1998). Cs extracts have high antioxidant activity,
primarily due to flavonoids and phenolic acids, and
in particular chlorogenic acid (5-caffeoylquinic acid),
dicaffeoylquinic acids and caffeic acid. Chlorogenic acid
is a potent inhibitor of glucose 6-phosphate translocase,
an essential component of the hepatic glucose 6-phos-
(23.6); 54.4 (2.9)
(25.8); 38.7 (3.5)
(10.1); 15.8 (2.4)
80.3 (6.2); 18.0 (1.3)
54.8 (4.9); 27.6 (1.8)

phatase system, which regulates the homeostasis of

intervention (dietary

15.2 (2.1); 7.6 (1.1)

blood glucose (Arion et al., 1997). Moreover, it has been

Last week of

supplement)

shown that, in rats, chlorogenic acid inhibits glucose ab-

7485 (563.1)

sorption from the small intestine (Welsch et al., 1989).

(32.5)
(3.6)

Furthermore, by inhibiting Glucose-6-Phospase (Glc-

6-Pase) activity, chlorogenic acid may limit the release
259.5
184.4
75.1
21.0
207.4

of glucose from glycogen into general circulation and pre-

vents the increase of insulinemia, as reported in animal
models, with the chlorogenic acid derivative (Herling
et al., 1998; Simon et al., 2000). Diabetes causes a two-
to threefold increase in Glc-6-Pase activity in the liver
(Segal and Washko, 1959; Arion and Nordlie, 1965),
making this enzyme system a potential target for com-
pounds that are intended to suppress hepatic glucose
236.1 (22.5); 54.6 (1.7)
160.9 (24.2); 37.8 (2.8)

production and to reduce hyperglycaemia. In addition,

49.7 (4.5); 27.7 (1.3)
70.9 (5.4); 17.5 (1.1)

72.6 (7.4); 16.9 (1.8)

intervention (dietary

13.3 (1.2); 7.4 (0.5)

dicaffeoylchinic acid derivatives may favour the glycaemic

First week of

supplement)

control by modulating the activity of alpha-glucosidase

and consequently the catabolism of dietary carbohydrates
6772.6 (554.8)

185.8 (31.1)

(Matsui et al., 2006).

23.9 (2.5)

The clinical impact that the results obtained in this

study could play is of relevant interest due to the high
prevalence in the western world of IFG (Shaw et al.,
2010) with increased risk of developing type 2 diabetes
(Unwin et al., 2002) and subsequently of developing
various CVDs.
The potential use of a Cs extract as a treatment of
IFG could represent an interesting alternative to the
use of glucose-lowering drugs avoiding the risk of un-
wanted symptoms and complications, such as severe
Saturated fatty acids (g);

hypoglycaemia, lactic acidosis, idiosyncratic liver cell

Protein (g); (% energy)

Carbohydrate (g); (%)

injury, permanent neurological deficit, digestive

Fat (g); (% energy)

Complex (g); (%)

discomfort, headache and dizziness (Neustadt and

Cholesterol (mg)
Dietary fiber (g)
Simple (g); (%)

Pieczenik, 2008; Arion and Nordlie, 1965; Modi,

(% energy)
Energy (kJ)

2007; Hui et al., 2005; Garber and Spann, 2008). As

a matter of fact, the intervention with Cs appears to
be safe: no relevant side effects were observed in
the intervention group.
Copyright © 2013 John Wiley & Sons, Ltd. Phytother. Res. 28: 33–41 (2014)
40 M. RONDANELLI ET AL.
Complementary medicine information needs to be
incorporated into clinical practice and patient and pro-
fessional education; awareness of the widespread use
of complementary and alternative medicine by people
with Type 2 diabetes is crucial for healthcare profes-
sionals (Thompson Coon and Ernst, 2003).
Another favourable effect observed in the present
study regards the lipidic pattern. After 8 weeks of
intervention with three daily doses of highly standar-
dized extract from Cs flowering heads, a significant re-
duction of TC, total/HDL-C ratio and LDL-C was
observed. The favourable effects of Cs supplementa-
tion on cholesterol metabolism, which are mostly
mediated by its choleretic activity, have been already
demonstrated in animal models (Saénz Rodriguez
et al., 2002) as well as in humans (Lupattelli et al.,
2004), so the results of this study confirm previous
findings.
It is important to consider that weight loss could also
have influenced the decrease of glucose and lipid values
because a significant effect on weight reduction has been
observed only in the intervention group, even though
both groups were asked to follow a similar prudent diet.
However, there was no significant correlation between
the decrease in fasting glycaemia and weight loss in trea-
ted group, which demonstrated that the weight loss is
not the cause of the decrease in fasting glycaemia.
Moreover, both groups consumed the same prudent
diet; probably, the more loss of weight observed in the
treatment group was due to a decrease in appetite
REFERENCES
American Diabetes Association and National Institute of Diabetes,
Digestive and Kidney Diseases. 2002. The prevention or delay
of type 2 diabetes. Diabetes Care 25: 742–749.
Arion WJ, Canfield WK, Ramos, . 1998. Chlorogenic acid analogue
S 3483: a potent competitive inhibitor of the hepatic and renal
glucose-6-phosphatase systems. Arch. Biochem. Biophys.
351: 279–285.
Arion WJ, Canfield WK, Ramos FC, . 1997. Chlorogenic acid and
hydroxynitrobenzaldehyde: new inhibitors of hepatic glucose
6-phosphatase. Arch. Biochem. Biophys. 339: 315–322.
Arion WJ, Nordlie RC. 1965. Liver glucose-6-phosphatase and
pyrophosphate-glucose phosphotransferase: effects of fast-
ing. Biochem. Biophys. Res. Commun. 20: 606–610.
Azzini E, Bugianesi R, Romano F, . 2007. Absorption and metabolism
of bioactive molecules after oral consumption of cooked edible
heads of Cynara scolymus L. (cultivar Violetto di Provenza) in
human subjects: a pilot study. Br. J. Nutr. 97:963–969.
DECODE Study Group, the European Diabetes Epidemiology
Group. 2001. Glucose tolerance and cardiovascular mortality:
comparison of fasting and 2-hour diagnostic criteria. Arch.
Intern. Med. 161: 397–405.
Dickel ML, Rates SM, Ritter MR. 2007. Plants popularly used for
loosing weight purposes in Porto Alegre, South Brazil. J. Eth-
nopharmacol. 9: 60–71.
Eastman RC, Cowie CC, Harris MI. 1997. Undiagnosed diabetes or
impaired glucose tolerance and cardiovascular risk. Diabetes
Care 20: 127–128.
Fantini N, Colombo G, Giori A, . 2010. Evidence of glycemia-
lowering effect by a Cynara scolymus L. extract in normal
and obese rats. Phytother. Res. 25: 463–466.
Frisancho AR. 1984. New standards of weight and body compos-
ition by frame size and height for assessment of nutritional sta-
tus of adults and the elderly. Am. J. Clin. Nutr. 40: 808–819.
Garber AJ, Spann SJ. 2008. An overview of incretin clinical trials.
J. Fam. Pract. 57: S10–S18.
Glümer C, Jǿrgensen T, Borch-Johnsen K. 2003. Prevalences of
diabetes and impaired glucose regulation in a Danish popula-
tion: the Inter99 study. Diabetes Care 26: 2335–2340.
Copyright © 2013 John Wiley & Sons, Ltd.
sensation. The losing weight effect of Cynara has been
already demonstrated both in clinical research, although
in combination with Phaseolus vugaris (Rondanelli
et al., 2011), and in review of traditional use of Cs
(Dickel et al., 2007).
These interesting results have been demonstrated in
naïve overweight patients, but it is assumed that the
same results can be evident in patients of normal weight,
seen that the weight loss did not affect blood glucose.
An interesting future study might involve association
to hypoglycaemic therapy with the intake of this dietary
supplement in severely ill patients with diabetes, with
aim to improve glucose metabolism.
Further studies with a larger number of patients are
clearly needed to confirm the effect of Cs extracts
on FBG.
Moreover, it will be important to prove this
hypoglycaemic effect in different study groups and not
just in overweight patients with newly detected IFG, as
it has been done in this study.
In conclusion, the supplementation with a highly stan-
dardized Cs extract is of great clinical interest for its po-
tential role in the treatment of overweight patients with
naïve IFG.
Conflict of Interest
All authors declare no financial/commercial conflicts of interest.
Haber GB, Heaton KW, Murphy D, Burroughs LF. 1977. Depletion
and disruption of dietary fibre. Effects on satiety, plasma-
glucose, and serum-insulin. Lancet 2: 679–682.
Haffner SM, Kennedy E, Gonzalez C, Stern MP, Miettinen H. 1996.
A prospective analysis of the HOMA model. The Mexico City
Diabetes Study. Diabetes Care 19: 1138–1141.
Hemmerle H, Burger HJ, Below P, . 1997. Chlorogenic acid and
synthetic chlorogenic acid derivatives: novel inhibitors of
hepatic glucose-6-phosphate translocase. J. Med. Chem.
40: 137–145.
Herling AW, Burger HJ, Schwab D, . 1998. Pharmacodynamic pro-
file of a novel inhibitor of the hepatic glucose-6-phosphatase
system. Am. J. Physiol. 274: G1087–G1093.
Hernandez-Galicia E, Aguilar-Contreras A, Aguilar-Santamaria L, .
2002. Studies on hypoglycemic activity of Mexican medicinal
plants. Proc. West. Pharmacol. Soc. 45: 118–124.
Hui H, Tang G, Go VL. 2009. Hypoglycemic herbs and their action
mechanisms. Chin Med 4: 11.
Hui H, Zhao X, Perfetti R. 2005. Structure and function studies of
glucagon-like peptide-1 (GLP-1): the designing of a novel
pharmacological agent for the treatment of diabetes. Diabetes
Metab. Res. Rev. 21: 313–331.
Karthikesan K, Pari L, Menon VP. 2010. Combined treatment of
tetrahydrocurcumin and chlorogenic acid exerts potential anti-
hyperglycemic effect on streptozotocin-nicotinamide-induced
diabetic rats. Gen. Physiol. Biophys. 29: 23–30.
Lupattelli G, Marchesi S, Lombardini R, . 2004. Artichoke juice
improves endothelial function in hyperlipemia. Life Sci.
76: 775–782.
Matsui T, Ogunwande IA, Abesundara KJ, Matsumoto K. 2006.
Anti-hyperglycemic potential of natural products. Mini Rev.
Med. Chem. 6: 349-356.
Modi P. 2007. Diabetes beyond insulin: review of new drugs for
treatment of diabetes mellitus. Curr. Drug Discov. Technol.
4: 39-47.
Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ;
A1c-Derived Average Glucose Study Group. 2008Translating
the A1C assay into estimated average glucose values.
Phytother. Res. 28: 33–41 (2014)
 CYNARA SCOLYMUS AND IMPAIRED GLYCAEMIA
Translating the A1C assay into estimated average glucose
values. Diabetes Care 31: 1473–1478.
Neustadt J, Pieczenik SR. 2008. Medication-induced mitochon-
drial damage and disease. Mol. Nutr. Food Res. 52: 780–788.
Rondanelli M, Giacosa A, Orsini F, Opizzi A, Villani S. 2011. Appe-
tite Control and Glycaemia Reduction in Overweight Subjects
treated with a Combination of Two Highly Standardized
Extracts from Phaseolus vulgaris and Cynara scolymus. Phyt-
other. Res. doi: 10.1002/ptr.3425.
Saénz Rodriguez T, Garcìa Giménez D, de la Puerta Vàzguez R.
2002. Choleretic activity and biliary elimination of lipids and
bile acids induced by an artichoke leaf extract in rats. Phyto-
medicine 9: 687–693.
Saydah SH, Loria CM, Eberhardt MS, Brancati FL. 2001. Subclin-
ical states of glucose intolerance and risk of death in the U.
S. Diabetes Care 24: 447–453.
Schütz K, Muks E, Carle R, Schieber A. 2006. Quantitative deter-
mination of phenolic compounds in artichoke-based dietary
supplements and pharmaceuticals by high-performance liquid
chromatography. J. Agric. Food Chem. 54: 8812–8817.
Segal HL, Washko ME. 1959. Studies of liver glucose 6-phosphat-
ase. III. Solubilization and properties of the enzyme from
normal and diabetic rats. J. Biol. Chem. 234: 1937–1941.
Shan JJ, Rodgers K, Lai CT, Sutherland SK. 2007. Challenges in
natural health product research: The importance of
standardization. Proc. West. Pharmacol. Soc. 50: 24–30.
Copyright © 2013 John Wiley & Sons, Ltd.
41
Shaw JE, Sicree RA, Zimmet PZ. 2010. Global estimates of the
prevalence of diabetes for 2010 and 2030. Diabetes Res. Clin.
Pract. 87: 4–14.
Simon C, Herling AW, Preibisch G, Burger HJ. 2000. Upregulation
of hepatic glucose 6-phosphatase gene expression in rats trea-
ted with an inhibitor of glucose-6-phosphate translocase.
Arch. Biochem. Biophys. 373: 418–428.
Steer RA, Cavalieri TA, Leonard DM, Beck AT. 1999. Use of the
Beck Depression Inventory for Primary Care to screen for
major depression disorders. Gen. Hosp. Psychiatry 21:
106–111.
Thompson Coon JS, Ernst E. 2003. Herbs for serum choles-
terol reduction: a systematic view. J. Fam. Pract. 52:
468–478.
Unwin N, Faughnan MS, Zimmet P, Alberti KG. 2002. Impaired
glucose tolerance and impaired fasting glycaemia: the
current status on definition and intervention. Diabet. Med.
19: 708–723.
Welsch CA, Lachance PA, Wasserman BP. 1989. Dietary phenolic
compounds: inhibition of Na + -dependent D-glucose uptake
in rat intestinal brush border membrane vesicles. J. Nutr.
119: 1698–1704.
World Health Organization. Energy and Protein Requirements Re-
port of a Joint FAO/WHO/UNU Expert Consultation. 1985.
WHO Technical Report Series No. 724 WHO: Geneva,
Switzerland.
Phytother. Res. 28: 33–41 (2014)