Clinical Nutrition ESPEN 22 (2017) 1e6

Contents lists available at ScienceDirect

Clinical Nutrition ESPEN

journal homepage: http://www.clinicalnutritionespen.com

Randomized Controlled Trial

Green tea extract outperforms metformin in lipid profile and

glycaemic control in overweight women: A double-blind,
placebo-controlled, randomized trial
Monallisa Alves Ferreira, Anna Paula Oliveira Gomes, Ana Paula Guimara ~es de Moraes,
~o Felipe Mota, Alexandre Siqueira Guedes Coelho,
Maria Luiza Ferreira Stringhini, Joa
Patrícia Borges Botelho*
Clinical and Sports Nutrition Research Laboratory (Labince), Faculty of Nutrition, Federal University of Goias, 74.605-080 Goiania, GO, Brazil

a r t i c l e i n f o s u m m a r y

Article history: Background & aims: Both green tea and metformin are used as adjuvants to treat and prevent compli-
Received 10 May 2017 cations associated with obesity; however, studies comparing their action and interaction in non-diabetic
Accepted 15 August 2017 overweight women have not been reported. Thus, the current study evaluated the effects of green tea
extract and metformin, both individually and in combination, on type 2 diabetes risk factors in non-
Keywords: diabetic overweight women.
Camellia sinensis
Methods: A total of 120 overweight women were randomly assigned in a double-blind manner to 1 of 4
Metformin
groups, as follows: control (n ¼ 29; 1 g of cellulose), green tea (n ¼ 32; 1 g of dry green tea extract),
Insulin resistance
Obesity
metformin (n ¼ 28; 1 g of metformin), and green tea þ metformin (n ¼ 31; 1 g of dry green tea
Lipoproteins LDL extract þ 1 g of metformin). Each group took the indicated capsules daily for 12 weeks. Anthropometric
measurements, body composition, and fasting blood samples were evaluated.
Results: Although no significant interactions were observed in glycaemic control (p ¼ 0.07), green tea in
the absence of metformin reduced fasting glucose (4.428 ± 2.00; p ¼ 0.031), but when combined the
lowering effect was nullified. In contrast, metformin increased HbA1c concentration (0.048 ± 0.189%;
p ¼ 0.017) and also reduced body weight (1.318 ± 0.366 kg; p ¼ 0.034) and LM (lean mass)
(1.249 ± 0.310; p ¼ 0.009). Regarding lipid parameters, green tea significantly reduced total cholesterol
and LDL-c.
Conclusions: Green tea was superior to metformin in improving glycaemic control and lipid profile in
non-diabetic overweight women and, therefore, green tea extract is a promising alternative for reducing
type 2 diabetes risk in overweight women.
© 2017 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights
reserved.

1. Introduction

Obesity is correlated with disturbances in pancreatic islets,

uncompensated glycaemic control and impaired glucose tolerance,
Abbreviations: LM, lean mass; FM, fat mass; PC, placebo group; GT, group all of which increase type 2 diabetes mellitus (DM2) risk [1]. Studies
receiving green tea extract supplementation alone; MF, group receiving metformin
show that lifestyle changes can delay or prevent DM2 in individuals
alone; GTMF, group receiving both green tea and metformin; Tea 0, groups
receiving only metformin and/or placebo; Tea 1, group receiving green tea extract
with obesity. In some cases, pharmacological interventions with
supplementation alone þ group receiving green tea extract and metformin com- oral antidiabetic agents such as metformin have been used to
bined; Met 0, groups receiving green tea extract alone and/or placebo; Met 1, group minimize DM2 risk even in non-diabetic obese patients [2]. How-
receiving metformin alone þ group receiving green tea extract and metformin ever, in a crossover study with non-diabetic obese women, met-
combined.
s, St. 227,
formin did not improve blood glucose as expected [3]. Also, lifestyle
* Corresponding author. Faculty of Nutrition, Federal University of Goia
rio, 74.605-080 Goia
Block 68, Setor Leste Universita ^nia, GO, Brazil. changes may be equal to or more effective than metformin in
E-mail address: patriciaborges.nutri@gmail.com (P. Borges Botelho). treating abnormal glucose levels [4].

http://dx.doi.org/10.1016/j.clnesp.2017.08.008
2405-4577/© 2017 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.

Downloaded for Anonymous User (n/a) at Diponegoro University from ClinicalKey.com by Elsevier on March 21, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.
2 M. Alves Ferreira et al. / Clinical Nutrition ESPEN 22 (2017) 1e6
Furthermore, side effects related to regular use of metformin,
such as diarrhoea and vomiting episodes, may lead to the discon-
tinuation of treatment [5]. Additionally, the cost-effectiveness of
long-term early pharmacologic treatment of overweight in-
dividuals who have not been diagnosed with pre-diabetes or DM2
is questionable [6].
Green tea possesses similar actions to metformin in controlling
glucose production and uptake, which may be useful for the pre-
vention and treatment of obesity-related complications [7]. Based
on its activity, green tea can improve glycaemic control [8], reduce
body fat [9] and ameliorate other risk factors for type 2 diabetes,
such as lipid profile [10].
To the best of our knowledge, studies comparing the effects of
metformin and green tea, both individually and in combination, on
body composition and DM2 risk factors in humans are not available.
Therefore, we conducted the current double-blind, placebo-
controlled, randomized trial to assess the effects of a green tea
extract and metformin, individually and in combination, on body
composition, lipid profile, and insulin resistance in non-diabetic
overweight women.
2. Subjects and methods
2.1. Materials
Green tea was analysed using high-performance thin-layer
chromatography (HPTLC), and encapsulated in a specialized labo-
ratory (Nathupharmus®, GO, Brazil). Each green tea capsule con-
tained 500 mg of Camellia sinensis leaf extract (extracted with
ethanol 80%), with 280 mg polyphenols, including catechins. Met-
formin was quantified using high-performance liquid chromatog-
raphy (HPLC), and colourimetric method was used to quantify
cellulose, which served as the placebo in our study. Each capsule
contained 500 mg of metformin or cellulose microcrystalline.
2.2. Subjects
Power calculation was performed based on an anticipated
change in glycated haemoglobin concentration mean (our primary
outcome). The estimated effect size was set to 1% decrease from the
values in the control group. With a variance of 1.5 (%)2, a power of
0.8 and a significance level of 0.05, an estimated 24 subjects in each
group, were needed. We estimated 25% of dropout rate and initi-
ated the study with 120 subjects.
Participants were recruited between June and August 2014
through assessments of clinical records from nutritional primary
care facilities and the endocrinology clinic within the Clinical
Hospital of Federal University of Goias, Brazil.
To assess this calculated population, a total of 191 women were
screened by a questionnaire containing the inclusion criteria for the
study: women aged 20e45 years with abnormal glucose concen-
trations evaluated by fasting blood glucose (>100 mg/dL) or gly-
cated haemoglobin (5.7%) and BMI > 28.9 kg/m2 or BMI > 27.5 kg/
m2 and HOMA-IR > 3.60 [11]. All these parameters were assessed in
the beginning of the study through biochemical tests.
The exclusion criteria were as follows: use of insulin or
glycaemia-lowering medication; any weight control treatment;
clinical diagnosis of diabetes, kidney disease, liver disease, heart
disease, hyperthyroidism or menopause; pregnancy or breast-
feeding; and daily consumption of any kind of tea.
2.3. Study design and diet intake
A 12-week randomized, double-blind, placebo-controlled study
was performed in accordance with the principles in the Declaration
Downloaded for Anonymous User (n/a) at Diponegoro University from ClinicalKey.com by Elsevier on March 21, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.
of Helsinki and Good Clinical Practice guidelines. Ethical approval
was obtained from the Ethics Committee of Federal University of
Goias (reference number 636.652). The trial was also registered at
ensaiosclinicos.gov.br (clinical trial RBR-4bdwxs). All participants
provided written informed consent to participate in this study after
being informed of its design both orally and in writing.
The participants were randomized into four groups using the
Unweighted Pair Group Method with Arithmetic Mean (UPGMA)
method as follows: Placebo (PC), 4 capsules containing microcrys-
talline cellulose/day (1 g/day); Green tea (GT), 2 capsules contain-
ing dry extract of green tea/day (1 g/day; 560 mg polyphenols) þ 2
capsules containing placebo/day; Metformin (MF), 2 capsules
containing metformin (1 g/day) þ 2 capsules containing placebo/
day; Green Tea þ Metformin (GTMF), 2 capsules containing dry
extract of green tea/day(1 g/day; 560 mg polyphenols) þ 2 capsules
containing metformin/day (1 g/day). The participants were
instructed to consume two capsules before breakfast and two
capsules before lunch for better absorption of both compounds and
nutrients. They were also instructed to maintain their normal
lifestyle habits, including those related to diet and physical exer-
cise, during the 12-week intervention.
All of the capsules had the same appearance. An independent
research group not involved in the study was responsible for
administering the supplements according to randomization pro-
tocol; thus, the investigators and participants were both blinded.
The blinding code was provided to the investigators after statistical
analysis was complete.
Adherence to treatment was assessed by counting the number
of capsules remaining when the participants returned to the lab-
oratory. To increase the compliance, all individuals received weekly
short messages on their cell phones and phone calls.
Compliance with the consumption of foods was monitored
through a 7-day food record questionnaire that included five
weekdays and two weekend days. The records were subsequently
evaluated by the research team to assure that the dietary infor-
mation was complete and accurate. Then, data on food records
were analysed using Nutriquanti® (2011, Sa ~o Paulo, SP, Brazil).
2.4. Primary and secondary outcome measures
The primary outcome measure in this study was glycated hae-
moglobin concentration, and the secondary outcome measures
were fasting glucose, lipid profile and body composition. We also
explored measures of pancreatic activity (HOMAs). All endpoints
were measured at baseline and at 12 weeks in all groups.
2.5. Side effects reported during the intervention
The side effects reported by the participants were classified into
different categories as follows: gastrointestinal, welfare, colour and
odour of urine, feeling of fullness, fluid intake, diarrhoea, nausea,
headache, heartburn, dry mouth, and other.
2.6. Anthropometric measurements and body composition
All measurements were performed by the same operator, who
followed standard procedures. Subject height was measured during
the selection phase to the nearest 0.5 cm with a stadiometer (Model
Standard, Sanny®). After voiding and while wearing light clothing,
each participant's body weight was measured to the nearest 0.1 kg
on a digital scale (Filizola®, Brazil). Waist circumference was
measured on undressed subjects at the midpoint between the
lower margin of the last palpable rib and the top of the iliac crest.
DXA assessments of fat mass (FM), lean mass (LM) and body fat
percentage were conducted using a General Electric Lunar Prodigy
 M. Alves Ferreira et al. / Clinical Nutrition ESPEN 22 (2017) 1e6
scanner (DPX NT, GE©) with Encore 2011 software (version 13.60,
GE Healthcare). The coefficients of variation (CVs) for the DXA tests
of LM and FM were 0.75% and 1.03%, respectively.
2.7. Biochemical analysis
Blood samples were collected from a peripheral vein the morning
after a 12-h fast. Total cholesterol, HDL-c, triglycerides and fasting
glucose were analysed using an enzymatic colourimetric method;
HbA1c by turbidimetry; and fasting insulin concentration by chem-
iluminescence (Miura 2000, Kovalent®, Sa ~o Paulo, Brazil). LDL-c and
very low-density lipoprotein cholesterol (VLDL-c) levels were
calculated according to the Friedewald et al. (1972) equation [12].
HOMA-IR and HOMA-b were calculated to evaluate insulin resistance
and functional capacity of pancreatic beta cells, respectively [13].
Renal and hepatic toxicity were monitored. Creatinine,
glutamic-oxaloacetic transaminase (GOT) and glutamic pyruvic
transaminase (GPT) were analysed using the colourimetric method
Full Stop (Architect c8000, Abbott Diagnostics®, Illinois, USA).
2.8. Statistics
Descriptive statistics (mean ± standard error) were calculated
for each variable in each group. Means were adjusted using the
method of least squares. The net change for the intervention was
calculated by the differences between the values recorded before
and after the intervention. Dietary variables were adjusted at each
stage of the study according to average calorie consumption using
the residual method [14].
The normality assumption of residuals was checked using the
Lilliefors test. All tests were performed using a critical significance
level of 5%. A 22 factorial design was applied to evaluate the sig-
nificance of each factor on biomarker response; the first factor (x1)
was the use of metformin, while green tea extract supplementation
was the second factor (x2), and the combination of both was ana-
lysed using two levels (0 ¼ absence and 1 ¼ presence). The sig-
nificance of the main effects (x1 and x2) as well as their interactions
(x1 vs. x2) were calculated using ANOVA, and the results were
expressed as bar charts. The bars represent the standardized effect
estimates (net change) corresponding to linear regression co-
efficients as follows: Yijkl ¼ m þ gi þ mj þ ck þ (mc)jk þ 3 ijkl, where
y ¼ response, m ¼ constant, gi ¼ fixed effect of the pairing group,
mj ¼ metformin's contribution to the response, ck ¼ green tea ex-
tract's contribution to the response, (mc)jk ¼ contribution of the
interaction between metformin and green tea extract to the
response, and 3 ijkl ¼ the error associated with the observations. All
statistical analyses were performed using R version 3.1.1.
3. Results
3.1. Subject characteristics
Initially, 191 women were enrolled: 153 subjects matched the
study inclusion criteria, and 120 agreed to participate in the study.
The subjects were randomly assigned to 1 of 4 intervention groups
(Fig. 1). Thirty-two women withdrew during the intervention: 14 for
personal reasons, 9 due to lack of contact, and 9 because of inade-
quate capsule intake. As a result, 88 participants completed the study
(Fig. 1). At baseline, all groups had homogeneous characteristics,
except for age in the GT group (p ¼ 0.049, Supplementary Table 1).
3.2. Energy and macronutrient intake
Energy and macronutrient intake are shown in Table 1. Protein
intake increased in the absence of isolated green tea (tea 0) and
Downloaded for Anonymous User (n/a) at Diponegoro University from ClinicalKey.com by Elsevier on March 21, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.
3
metformin (met 0); however, no differences were observed when
interaction between treatments was analysed.
3.3. Body composition
Green tea supplementation did not affect body composition
outcomes (Table 2). In contrast, metformin reduced weight
(1.318 ± 0.366; p ¼ 0.034), BMI (0.460 ± 0.153; p ¼ 0.016) and
LM (1.499 ± 0.386; p ¼ 0.02) (Table 2).
3.4. Glycaemic control
Although no significant interactions were observed in our study
(p ¼ 0.07), green tea's effects under metformin absence and pres-
ence seemed to be different. Fig. 2 shows that green tea reduced
fasting glucose (4.428 ± 2.00; p ¼ 0.031) but when combined to
metformin the lowering effect was not observed. Moreover, met-
formin also increased glycated haemoglobin concentration
(0.048 ± 0.18; p ¼ 0.017). Thus, this difference of behaviour can
justify the trend about the interaction. Neither intervention
affected insulin, HOMA-b or HOMA-IR.
3.5. Lipid profile
Metformin did not change lipid profiles; however, significant
influence of green tea was observed in the reduction of total
cholesterol (11.467 mg/dL ± 3.71; p ¼ 0.047) and LDL-c
(5.031 mg/dL ± 3.82; p ¼ 0.024) (Table 2).
3.6. Hepatic and renal toxicity
Neither green tea nor metformin caused acute liver or kidney
toxicity. Metformin decreased creatinine concentrations at the end
of the study (0.023 mg/dL ± 0.01; p ¼ 0.009). Moreover, this effect
was also attenuated by the combination of green tea and metformin
(0.049 mg/dL ± 0.01; p ¼ 0.000) (Table 2).
3.7. Signs and symptoms reported by subjects during the
intervention
The symptoms among the subjects, including the placebo group,
were intestinal upset and the feeling of being full. In the metformin
group, episodes of diarrhoea (17%) and nausea (9%) were frequent.
When both treatments were combined, these side effects were
increased (26% and 16% for diarrhoea and nausea, respectively)
(Supplementary Table 2).
4. Discussion
To the best of our knowledge, the current study is the first
double-blind, placebo-controlled, randomized trial to investigate
the effects of green tea and metformin, individually and in com-
bination, on lipid profile, glycaemic control and body composition
in overweight humans, all of which are considered risk factors for
insulin resistance and DM2.
We found that 12 weeks of green tea treatment reduced total
cholesterol, LDL-c and fasting glucose levels in metformin absence.
In contrast, metformin increased glycated haemoglobin levels and
reduced body weight due to LM loss.
Both green tea and metformin are thought to be attractive
molecular triggers to the AMP-activated protein kinase (AMPK)
pathway, which is related to the translocation of GLUT-4 and the
improvement of glucose uptake into the intracellular medium, in-
hibition of hepatic gluconeogenesis through a decreased expres-
sion of gluconeogenic genes, attenuation in oxidative stress and the
4 M. Alves Ferreira et al. / Clinical Nutrition ESPEN 22 (2017) 1e6

Fig. 1. Schematic of participants flow through the study design.

Table 1
Least squares means estimates for green tea, metformin and their combination on energy and macronutrient intake.

Variable Green tea Metformin p Interaction

tea 0 tea 1 p met 0 met 1 p

Calories (kcal) 40.110 (105.61) 85.344 (119.82) 0.883 18.886 (101.55) 144.340 (120.71) 0.266 0.096
Protein (g) 19.342 (4.98) 0.235 (5.71) 0.016 18.112 (4.78) 1.464 (4.72) 0.019 0.310
Lipids (g) 1.562 (2.28) 2.343 (2.63) 0.222 0.997 (2.19) 1.779 (2.62) 0.356 0.397
Carbohydrates (g) 20.565 (7.50) 1.195 (8.39) 0.082 20.326 (7.10) 1.434 (8.50) 0.068 0.532
Fibre (g) 2.529 (1.27) 4.713 (1.43) 0.246 2.488 (1.22) 4.754 (1.44) 0.213 0.192

Data in bold refers to p-value < 0.05.

Data expressed as D ± standard error (se). Tea 0: groups receiving only metformin and/or placebo. Tea 1: group receiving green tea extract supplementation alone þ group
receiving green tea extract and metformin combined. Met 0: groups receiving green tea extract alone and/or placebo. Met 1: group receiving metformin alone þ group
receiving green tea extract and metformin combined. p-Interaction: p-value for interaction when there was combined use of the products.

increase fatty acid oxidation [15]. However, Zang et al. [16] sug-
gested that in animal models the effect of green tea polyphenols is
50e200 times more potent than metformin in AMPK activation.
This may justify the superior effect of green tea in reducing the
glycaemia in our study.
Contrary to our expectations, metformin did not reduce the
glycaemia. It increased glycated haemoglobin and still seemed to
nullify the green tea effect on fasting glucose. DeFronzo and Abdul-
Ghani [17] suggested that metformin initially reduces HbA1c levels
but this is followed by a progressive increase. Moreover, some

Downloaded for Anonymous User (n/a) at Diponegoro University from ClinicalKey.com by Elsevier on March 21, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.
 M. Alves Ferreira et al. / Clinical Nutrition ESPEN 22 (2017) 1e6 5

Table 2
Least squares means estimates for green tea, metformin and their combination on type 2 diabetes risk factors.

Variable Green tea Metformin Interaction p Interaction

tea 0 tea 1 p met 0 met 1 p

Weight (kg) 1.167 (0.34) 0.463 (0.35) 0.096 0.312 (0.33) 1.318 (0.37) 0.034 1.081 (0.51) 0.632
Waist circumference (cm) 0.959 (0.48) 1.138 (0.52) 0.881 0.705 (0.48) 1.391 (0.53) 0.347 1.985 (0.74) 0.155
BMI (kg/m2) 0.349 (0.14) 0.086 (0.15) 0.136 0.025 (0.14) 0.460 (0.15) 0.016 0.386 (0.21) 0.573
LM (kg) 1.086 (0.29) 0.856 (0.29) 0.561 0.443 (0.28) 1.499 (0.31) 0.026 1.134 (0.43) 0.310
FM (kg) 0.214 (0.33) 0.096 (0.34) 0.450 0.044 (0.32) 0.163 (0.35) 0.718 0.413 (0.50) 0.105
Fasting glucose (mg/dL) 0.839 (1.47) 2.289 (1.54) 0.085 0.952 (1.40) 0.499 (1.63) 0.935 0.151 (2.30) 0.070
Glycated haemoglobin (%) 0.317 (0.17) 0.152 (0.18) 0.615 0.516 (0.16) 0.048 (0.18) 0.017 0.033 (0.26) 0.181
Insulin (mg/dL) 1.104 (0.88) 0.731 (0.88) 0.750 0.812 (0.83) 1.024 (0.92) 0.839 0.572 (1.28) 0.657
iHOMA-IR 0.193 (0.20) 0.240 (0.21) 0.835 0.196 (0.19) 0.237 (0.22) 0.828 0.154 (0.32) 0.457
iHOMA b 5.853 (16.36) 9.441 (17.80) 0.452 6.799 (16.05) 10.387 (18.23) 0.360 4.763 (25.81) 0.054
Total cholesterol (mg/dL) 1.545 (3.49) 11.467 (3.71) 0.047 6.534 (3.46) 6.478 (3.72) 0.936 13.882 (5.25) 0.327
HDL-c (mg/dL) 3.570 (1.00) 5.187 (1.05) 0.218 4.981 (0.99) 3.776 (1.07) 0.414 3.772 (1.49) 0.251
LDL-c (mg/dL) 5.987 (3.66) 5.031 (3.82) 0.024 2.669 (3.53) 3.626 (3.89) 0.205 2.375 (5.49) 0.847
VLDL-c (mg/dL) 4.665 (1.40) 1.885 (1.45) 0.151 2.916 (1.35) 3.634 (1.50) 0.706 1.908 (2.09) 0.733
Triglycerides (mg/dL) 23.341 (6.95) 9.274 (7.20) 0.143 14.531 (6.70) 18.085 (7.46) 0.706 9.321 (10.38) 0.723
GOT (U/L) 0.268 (0.71) 0.320 (0.74) 0.979 0.081 (0.71) 0.506 (0.75) 0.656 0.861 (1.03) 0.520
GOT (U/L) 2.475 (1.00) 2.516 (1.05) 0.995 1.963 (0.96) 3.027 (1.08) 0.432 2.991 (1.48) 0.935
Creatinine (mg/dL) 0.005 (0.01) 0.009 (0.01) 0.957 0.009 (0.01) 0.023 (0.01) 0.009 0.049(0.01) 0.000

Data in bold refers to p-value < 0.05.

Data expressed as D ± standard error (se). Tea 0: groups receiving only metformin and/or placebo. Tea 1: group receiving green tea extract supplementation alone þ group
receiving green tea extract and metformin combined. Met 0: groups receiving green tea extract alone and/or placebo. Met 1: group receiving metformin alone þ group
receiving green tea extract and metformin combined. p-Interaction: p-value for interaction when there was combined use of the products. BMI: body mass index; LM: Lean
Mass; FM: Fat mass; iHOMA-IR: Homeostasis Model Assessment-Insulin Resistance index; iHoma-b: Homeostasis Model Assessment-b index; HDL: high-density lipoprotein;
LDL: low-density lipoprotein; VLDL: very low-density lipoprotein; GOT: glutamic oxaloacetic transaminase; TGP: glutamic pyruvic transaminase.

Fig. 2. Influence of green tea and metformin on fasting glucose.

studies have shown an estimate of 35% failure rate for metformin
monotherapy [18]. Thus, the variabilities in the response of met-
formin may explain the lack of effect of the drug in glucose con-
centration in the current study. Studies with animal models have
shown that metformin can fail to activate AMPK in the liver and
fails to induce a glucose-lowering effect due to the alterations in the
expression of genes involved in hepatic gluconeogenesis and lipo-
genesis [19], or even in metformin transporters [20]. Thus, further
studies in the field of pharmacogenomics are necessary to explain
these results. However, hypothesis involves metformin influencing
genic and intergenic regions [18].
One of our hypotheses was that green tea would reduce gly-
caemia through changes in body composition. However, we did not
find a difference in fat mass between groups. Hursel et al. [21]
suggest that epigallocatechin gallate (EGCG), a polyphenol found
in green tea, may be more effective in Asians because this popu-
lation appears to carry a polymorphism in the catechol O-methyl-
transferase (COMT) allele that increases COMT enzyme activity and
so adrenergic-induced lipolysis. Moreover, Asians also present
greater abdominal fat percentages and lower insulin sensitivity
than Caucasians [22].
In contrast, metformin induced weight loss, which corroborates
the findings in other studies in non-diabetic obese individuals
[23,24]. However, in this study, metformin reduced LM and did not
change FM. A possible relationship between metformin and muscle
mass reduction was demonstrated in a culture of muscle cells in
which the association between metformin and the expression of
the Muscle RING-finger gene protein-1 (MuRF1) was observed. This

Downloaded for Anonymous User (n/a) at Diponegoro University from ClinicalKey.com by Elsevier on March 21, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.
6 M. Alves Ferreira et al. / Clinical Nutrition ESPEN 22 (2017) 1e6
protein belongs to the family of E3 ubiquitin ligases and it induces
autophagic cell death [25].
Although no change has been found for body composition, green
tea decreased total cholesterol and LDL-c levels. Similar results were
observed in a meta-analysis of 14 studies of nondiabetic adults [26].
The reported mechanisms of action to explain this effect include
reduction in intestinal cholesterol absorption [27], a decrease in
hepatic production of cholesterol [28] and the modulation of LDL-c
receptors in the liver [29]. It is also suggested that AMPK activation
through green tea inhibits adipocyte differentiation, suppresses the
expression of lipogenic molecules and induces fatty acid oxidation
by decreasing the concentration of malonyl-CoA [30].
In summary, 12 weeks of green tea treatment ameliorated the
lipid profile and glycaemic control in the study subjects. In contrast,
the isolated metformin was not positive for glycaemic control and
influenced LM loss. Therefore, green tea may be a better alternative
to reduce metabolic damages, the risk of insulin resistance and
DM2 in overweight women than metformin. The limitations of the
current trial include not controlling for caffeine and milk intake, or
ethnicity.
Conflict of interest
The authors declare no conflicts of interest.
The authors' responsibilities were as follows: MAF and PBB
participated in study conception, study design, and drafting of the
manuscript; ASGC performed statistical analysis; APOG and FM
performed laboratory measurements; ANM and MLFS engaged in
participant recruitment. All authors participated in data interpre-
tation and critical review and revision of the manuscript, and all
approved the final manuscript.
Acknowledgements
We thank the Fundaça ~o de Amparo a  Pesquisa do Estado de
s (FAPEG) and Coordenaça
Goia ~o de Aperfeiçoamento de Pessoal de
Nível Superior (CAPES) for scholarships as well as Nathupharmu-
s®Pharmacyfor providing the intervention material.
Appendix A. Supplementary material
Supplementary material associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.clnesp.
2017.08.008.
References
[1] Grundy SM. Pre-diabetes, metabolic syndrome, and cardiovascular risk. J Am
Coll Cardiol 2012;59:635e43.
[2] Salpeter SR, Nicholas SB, Kahn JA, Salpeter EE. Meta-analysis: metformin
treatment in persons at risk for diabetes mellitus. Am J Med 2008;121:
149e57.
[3] Binnert C, Seematter G, Tappy L, Giusti V. Effect of metformin on insulin
sensitivity and insulin secretion in female obese patients with normal glucose
tolerance. Diabetes Metab 2003;29(2 Pt 1):125e32.
[4] Diabetes Prevention Program Research Group. Reduction in the incidence of
type 2 diabetes with lifestyle intervention or metformin. N Engl J Med
2002;346:393e403.
Downloaded for Anonymous User (n/a) at Diponegoro University from ClinicalKey.com by Elsevier on March 21, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.
[5] Bouchoucha M, Uzzan B, Cohen R. Metformin and digestive disorders. Dia-
betes Metab 2011;37:90e6.
[6] Levri KM, Slaymaker E, Last A, Yeh J, Ference J, D'Amico F, et al. Metformin as
treatment for overweight and obese adults: a systematic review. Ann Fam
Med 2005;3:457e61.
[7] Park CS, Bang BR, Kwon HS, Moon KA, Kim TB, Lee KY, et al. Metformin re-
duces airway inflammation and remodeling via activation of AMP-activated
protein kinase. Biochem Pharmacol 2012;84:1660e70.
[8] Takahashi M, Miyashita M, Suzuki K, Bae SR, Kim HK, Wakisaka T, et al. Acute
ingestion of catechin-rich green tea improves postprandial glucose status and
increases serum thioredoxin concentrations in postmenopausal women. Br J
Nutr 2014;112:1542e50.
[9] Westerterp-Plantenga MS. Green tea catechins caffeine and body-weight
regulation. Physiol Behav 2010;100:42e6.
[10] Onakpoya I, Spencer E, Heneghan C, Thompson M. The effect of green tea
on blood pressure and lipid profile: a systematic review and meta-analysis
of randomized clinical trials. Nutr Metab Cardiovasc Dis 2014;24:823e36.
[11] Stern SE, Williams K, Ferrannini E. Identification of individuals with insulin
resistance using routine clinical measurements. Diabetes 2005;54:333e9.
[12] Friedewald WT, Levi RI, Fredrickson DS. Estimation of the concentration of
low density lipoproteins cholesterol in plasma without use of the ultracen-
trifuge. Clin Chem 1972;18:499e502.
[13] Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC.
Homeostasis model assessment: insulin resistance and beta-cell function from
fasting plasma glucose and insulin concentrations in man. Diabetologia
1985;28:412e9.
[14] Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in
epidemiologic studies. Am J Clin Nutr 1997;65:1220e8.
[15] Hardie DG, Ross FA, Hawley SA. AMPK: a nutrient and energy sensor that
maintains energy homeostasis. Nat Rev Mol Cell Biol 2012;13:251e62.
[16] Zang M, Xu S, Maitland-Toolan KA, Zuccollo A, Hou X, Jiang B, et al. Poly-
phenols stimulate AMP-activated protein kinase, lower lipids, and inhibit
accelerated atherosclerosis in diabetic LDL receptor-deficient mice. Diabetes
2006;55:2180e91.
[17] DeFronzo RA, Abdul-Ghani MA. Preservation of b-cell function: the key to
diabetes prevention. J Clin Endocrinol Metab 2011;96:2354e66.
[18] Pawlyk AC, Giacomini KM, McKeon C, Shuldiner AR, Florez JC. Metformin
pharmacogenomics: current status and future directions. Diabetes 2014;63:
2590e9.
[19] Shaw RJ, Lamia KA, Vasquez D, Koo SH, Bardeesy N, Depinho RA, et al. The
kinase LKB1 mediates glucose homeostasis in liver and therapeutic effects of
metformin. Science 2005;310:1642e6.
[20] Holstein A, Seeringer A, Kovacs P. Therapy with oral antidiabetic drugs:
applied pharmacogenetics. Br J Diabetes Vasc Dis 2011;11:10e6.
[21] Hursel R, Viechtbauer W, Westerterp-Plantenga MS. The effects of green tea
on weight loss and weight maintenance: a meta-analysis. Int J Obes (Lond)
2009;33:956e61.
[22] Wulan SN, Westerterp KR, Plasqui G. Ethnic differences in body composition
and the associated metabolic profile: a comparative study between Asians and
Caucasians. Maturitas 2010;65:315e9.
[23] Glueck CJ, Fontaine RN, Wang P, Subbiah MT, Weber K, Illiq E, et al. Metformin
reduces weight, centripetal obesity, insulin, leptin, and low-density lipopro-
tein cholesterol in nondiabetic, morbidly obese subjects with body mass index
greater than 30. Metab Clin Exp 2001;50:856e61.
[24] Golay A. Metformin and body weight. Int J Obes 2008;32:61e72.
[25] Krawiec BJ, Nystrom GJ, Frost RA, Jefferson LS, Lang CH. AMP activated
protein kinase agonists increase mRNA content of the muscle-specific
ubiquitin ligases MAFbx and MuRF1 in C2C12 cells. Am J Physiol Endo-
crinol Metab 2007;292:1555e67.
[26] Zheng XX, Xu YL, Li SH, Liu XX, Hui R, Huang XH. Green tea intake lowers
fasting serum total and LDL cholesterol in adults: a meta-analysis of 14 ran-
domized controlled trials. Am J Clin Nutr 2011;94:601e10.
[27] Koo SI, Noh SK. Green tea as inhibitor of the intestinal absorption of lipids:
potential mechanism for its lipid-lowering effect. J Nutr Biochem 2007;18:
179e83.
[28] Kim A, Chiu A, Barone MK, Avino D, Wang F, Coleman CI, et al. Green tea
catechins decrease total and low-density lipoprotein cholesterol: a systematic
review and meta-analysis. J Am Diet Assoc 2011;111:1720e9.
[29] Bursill CA, Roach PD. A green tea catechin extract upregulates the hepatic
low-density lipoprotein receptor in rats. Lipids 2007;42:621e7.
[30] Huang J, Wang Y, Xie Z, Zhou Y, Zhang Y, Wan X. The anti-obesity effects of
green tea in human intervention and basic molecular studies. Eur J Clin Nutr
2014;68:1075e87.