Clinical Nutrition ESPEN 10 (2015) e95ee101

Contents lists available at ScienceDirect

Clinical Nutrition ESPEN

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

Original article

Chlorella ingestion suppresses resistin gene expression in peripheral

blood cells of borderline diabetics
Hiroshige Itakura a, Michie Kobayashi b, *, Seiji Nakamura b
a
Shinagawa East One Medical Clinic, East One Tower 3F, 2-16-1, Konan, Minato-Ku, Tokyo, Japan
b
DNA Chip Research Inc., 1-1-43, Suehiro-cho, Tsurumi-ku, Yokohama, Japan

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

Article history: Background & aims: Type 2 diabetes can lead to arteriosclerosis, renal damage, retinopathy, and several
Received 4 November 2014 other serious complications. To prevent or delay the progression of this disease, considerable attention
Accepted 1 April 2015 has been paid to improve exercise and dietary habits. Chlorella ingestion can reportedly reduce high
blood glucose and cholesterol levels in mice and humans, although no studies have critically evaluated
Keywords: the effects of Chlorella on human borderline diabetics. Thus, we conducted a randomized, double-blind
Resistin
placebo-controlled test with volunteer borderline diabetics.
Borderline-diabetes
Methods: We recruited 57 subjects and randomly divided into a Chlorella ingestion group (n ¼ 28) and a
Chlorella
Gene-expression
Placebo ingestion group (n ¼ 29). Blood samples were collected every 4 weeks for laboratory tests. Gene
Microarray expression analyses using peripheral blood cell RNA were performed before and 12 weeks after the trial.
Results: A total of 252 genes showed changed expression levels between these two groups. Six of these
were type 2 diabetes-associated genes, including resistin, an insulin resistance inducer that exhibited
markedly reduced expression with Chlorella ingestion (P ¼ 0.01). Resistin mRNA expression significantly
correlated with changes in HbA1c and TNF-a and IL-6 levels, all of which are strongly associated with
glucose metabolism and/or inflammation.
Conclusions: Chlorella ingestion may be useful in preventing or ameliorating the course of type 2 dia-
betes development. In addition, gene expression analysis may be a means to investigate the effects of
foods and supplements in humans.
© 2015 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights
reserved.

1. Introduction
Type 2 diabetes incidence is increasing worldwide. This disease
can lead to the development of various complications, such as
arteriosclerosis, renal failure [1,2], and retinopathy, all of which
Abbreviations: HDL, high-density lipoprotein; VLDL, very-low-density lipopro-
tein; LDL, low-density lipoprotein; HbA1c, hemoglobin A1c; FOSHU, food for
specified health uses; GA, glycosylated albumin; 1,5-AG, 1,5-anhydroglucitol; Apo-
B, cholesterol and apolipoprotein B; IL, Interleukin; TNF, -6,tumor necrosis factor;
PAI-1, -aadiponectin and plasminogen activator inhibitor-1; ELISA, enzyme-linked
immunosorbent assay; HPLC, high-performance liquid chromatography; QRT-PCR,
quantitative reverse-transcription polymerase chain reaction; BMI, body mass in-
dex; ADIPOR1, adiponectin receptor 1; AMPK, activated protein kinase; LDLR, LDL
receptor; PER1, period homolog 1; PPARD, peroxime proliferator-activated receptor
delta; ALMS1, alstrom syndrome 1; AUTS2, autism susceptibility candidate 2.
* Corresponding author. DNA Chip Research Inc., Research & Development Group,
1-1-43, Suehiro-cho, Tsurumi-ku, Yokohama, Japan. Tel.: þ81 45 500 5218; fax: þ81
45 500 5219.
E-mail address: kobayashi@dna-chip.co.jp (M. Kobayashi).
affect the quality of life of patients [3]. Several borderline (pre-
diabetic) status patients, who want to avoid these adverse com-
plications [4], engage in exercise and change their dietary habits to
reduce the risk of type 2 diabetes [5,6]. In addition, their needs for
appropriate supplement intake are high. Chlorella ingestion can
reportedly improve glucose intolerance and dyslipidemia [7e9].
Chlorella is a unicellular green alga that can be ingested as a
powder or a hot water extract. This food contains high concentra-
tions of natural-origin essential amino acids, vitamins, and minerals.
The effects of ingesting Chlorella have been investigated in animal
models, such as obese model mice (ob/ob mice) that showed sig-
nificant reductions in serum total cholesterol levels and upregulated
insulin concentrations with concomitant increases in adiponectin
concentrations [7] or diabetic rat models that showed favorable ef-
fects on high-density lipoprotein (HDL), very low-density lipopro-
tein (VLDL), and low-density lipoprotein (LDL) cholesterol
concentrations [9]. In humans with possible “lifestyle-related dis-
eases,” Chlorella ingestion resulted in improved body fat percent-
ages, serum total cholesterol concentrations, and fasting blood

http://dx.doi.org/10.1016/j.clnesp.2015.04.002
2405-4577/© 2015 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.
e96 H. Itakura et al. / Clinical Nutrition ESPEN 10 (2015) e95ee101

glucose concentrations [8]. Another report showed improvements 2.2. Clinical laboratory tests
in insulin resistance at the gene level for diabetic rats [9].
However, few studies have systematically investigated the ef- Serum was obtained by centrifuging coagulated blood samples.
fects of Chlorella ingestion in humans. In this study, we report on HbA1c concentrations were determined by a latex aggregation
our results of using Chlorella ingestion based on a randomized, method. Serum interleukin (IL)-6, rumor necrosis factor (TNF)-a,
double-blind, parallel, placebo-controlled trial for its blood-borne and adiponectin levels were determined by enzyme-linked
effects. The subjects of our study were borderline diabetics. If immunosorbent assay (ELISA). Total homocysteine levels were
Chlorella intake would be able to suppress progression of glucose determined using high-performance liquid chromatography
intolerance, pre-diabetic persons might gain beneficial quality of (HPLC).
life. Chlorella used in this study was pulverized Chlorella pyr-
enoidosa (Sun Chlorella strain). We performed comprehensive gene 2.3. Total RNA extraction from blood
expression analyses using peripheral blood cells along with records
of physiological changes. Peripheral blood RNA reflects the tran- A peripheral blood sample was collected in a PAXgene blood
scriptome activities of circulating immune cells and can be used as RNA tube (BD, Franklin Lakes, NJ, USA), left at room temperature for
a barometer of health conditions, such as fatigue [10], obesity [11], 2 h after sufficient mixing and subsequently stored at 20  C. The
and type 2 diabetes [12]. In addition, only a few studies have re- stored tube was used for total RNA extraction using a PAXgene
ported on these effects associated with foods [13,14]. We used Blood miRNA Kit (QIAGEN, Valencia, CA, USA).
blood cell RNA assays to determine any subtle effects that may have The amount of extracted total RNA was determined using a
occurred during a long trial period. NanoDrop 1000 (Thermo Fisher Scientific Inc., Waltham, MA). RNA
Our results showed that Chlorella ingestion affected resistin degradation was checked using an Agilent 2100 Bioanalyzer (Agi-
gene expression, an indicator of ameliorating type 2 diabetes, and lent Technologies, Palo Alto, CA, USA).
that this effect paralleled reductions in HbA1c, IL-6, and TNF-a
levels. These results strongly suggested that inflammation was also 2.4. Gene expression microarray analysis
affected for ameliorating this condition. Furthermore, resistin gene
expression negatively correlated with the adiponectin receptor Microarray analyses were performed with SurePrint G3 human
gene expression levels, which again reflected favorable effects of GE 8  60K microarrays (Design ID; 028004, Agilent Technologies).
Chlorella ingestion on ameliorating type 2 diabetes development. Cyanine 3-labeled cRNA was prepared from 100 ng of total RNA
using a Low Input Quick Amp Labeling Kit (Agilent Technologies)
2. Methods according to the manufacturer's protocol. Microarrays were scan-
ned using a G2505C DNA Microarray Scanner (Agilent Technolo-
2.1. Subjects gies). The images were converted to numerical values by Feature
Extraction Software ver. 10.7.3.1 (Agilent Technologies).
The subjects were adult Japanese men aged 40e54 years and Data analysis was performed using GeneSpring GX v11.5 (Agi-
who had borderline diabetes as defined by their hemoglobin A1c lent Technologies), with which inter-array normalization was per-
(HbA1c). Their HbA1c (NGSP values) concentrations were between formed using the 75th percentile, and inter-gene normalization
6.0% and 6.5%. We excluded subjects who had any history of serious was performed on the basis of the median value of all samples.
diseases of the metabolic and/or endocrine systems, Food for There are total of 42,405 probes on Agilent Human GE 8  60K
Specified Health Uses (FOSHU), and any allergy to foods and/or Microarray (Design ID: 028004) without control probes. Signals
drugs. We also excluded those who had received any medications. were categorized as either present, marginal, or absent based on
Date of approval of the ethics committee is June 17, 2010. We car- the signal of each probe after conversion to a numerical value for
ried out recruitment and the follow-up of the patient in a period
from June 18, 2010 to December 18, 2010. Subjects were randomly Table 1
assigned following simple randomization procedures to 1 of 2 Subjects' clinical characteristics.a
treatment groups (30/group): a Chlorella group that ingested bulk
Variable Chlorella group Placebo group
Chlorella (8.0 g/day) and a Placebo group that ingested a lactose
formulation (8.0 g/day). Subjects were asked to participate in this N 28 29
Age (years) 46.4 ± 3.6 46.9 ± 3.9
test for 12 weeks. The placebo was matched to the study food for BMI (kg/m2) 26.5 ± 0.8 25.6 ± 0.6
taste, color, and size. The subjects were monitored for 4 weeks Abdominal girth (cm) 93.1 ± 2.03 89.2 ± 1.7
following these administrations. The Chlorella powder used was Triglyceride (mg/dL) 148.5 ± 19.8 179.1 ± 22.9
“Sun Chlorella A” tablets (Sun Chlorella, Corp., Kyoto, Japan). Overall Blood pressure (max; mmHg) 124.0 ± 2.4 123.5 ± 2.3
Blood pressure (mini; mmHg) 82.0 ± 1.9 81.6 ± 2.1
subject compliance with food ingestion was estimated by food
Total cholesterol (mg/dL) 220.3 ± 6.9 221.0 ± 6.5
diaries. HDL cholesterolb (mg/dL) 56.7 ± 3.3 55.5 ± 2.5
Blood, including HbA1c levels, and urine analyses were per- LDL cholesterolb (mg/dL) 130.3 ± 7.1 128.5 ± 5.7
formed at 0, 4, 8, and 12 weeks during the test period and at 4 Apo-Bb (mg/dL) 103.3 ± 3.2 106.2 ± 3.2
weeks after the test period. We collected blood under non-fasting Blood glucose (mg/dL) 97.6 ± 1.9 95.2 ± 1.4
1.5-AGb (%) 18.9 ± 1.1 21.5 ± 1.4
conditions. The subjects in these groups were not significantly HbA1cb(%) 6.0 ± 0.0 6.1 ± 0.0
different with regard to their demographic and clinical GAb (%) 14.2 ± 0.2 14.3 ± 0.2
backgrounds. IL-6b (pg/mL) 1.6 ± 0.1 1.5 ± 0.2
This study conformed to the Declaration of Helsinki and the CRPb (mg/mL) 119.4 ± 24.8 117.3 ± 23.2
Adiponectin (mg/mL) 8.7 ± 0.7 8.4 ± 0.5
Ethical Guidelines for Epidemiological Studies. Protection of human
TNF-ab (pg/mL) 0.7 ± 0.1 0.6 ± 0.1
rights of the subjects was always considered, and our study pro-
a
tocol was approved by the Institutional Review Board of the Values are means ± S.E.M (Chlorella vs. Placebo) (t-test: #P  0.05).
b
Apo-B, Apolipoprotein B; 1.5-AG, 1.5-Anhydro-D-glucitol; HbA1c, Hemoglobin
Chiyoda Paramedical Care Clinic. Written informed consent was A1c (NGSP); GA, Glycoalbumin; HDL cholesterol, High density lipoprotein choles-
obtained from all participants. This study is registered in the UMIN terol; LDL cholesterol, Low density lipoprotein cholesterol; IL-6, Interleukin-6; CRP,
Clinical Trials Registry with the identifier UMIN000009130. C-reactive protein; TNF-a, Tumor necrosis factor-alpha.
 H. Itakura et al. / Clinical Nutrition ESPEN 10 (2015) e95ee101 e97

further analyses. Those partials with signals present (Flag “P”;
Present) were used. A total of 18,535 probes that satisfied the
condition of >80% Flag “P” in both the Chlorella and Placebo groups
were first considered for analysis. We subsequently excluded 1052
probes that showed significant differences in signal intensities
between the two groups at week 0 (Welch's t-test, P  0.05;
without multiple testing corrections), and used the remaining
17,511 probe set for further analyses.
2.5. Analysis A (inter-group comparisons)
The extent of changes in gene expressions between weeks 0 and
12 were determined in each study group and subsequently
compared by Welch's t-test (P  0.05).
2.6. Analysis B (intra-group comparisons)
registered in a public database, NCBI Gene Expression Omnibus
(GEO; http://www.ncbi.nlm.nih.gov/geo), with the GEO registra-
tion numbers GSE41767.
2.7. Quantitative reverse-transcription polymerase chain reaction
(QRT-PCR)
cDNA was synthesized from 1000 ng of extracted total RNA
using a High Capacity cDNA Reverse Transcription Kit (Applied
Biosystems, Foster City, CA). Real-time PCR was performed in
triplicate using an ABI 7500 Real-Time PCR System (Applied Bio-
systems). Primers and TaqMan probes for the target gene (resistin:
RETN) and an internal reference gene (beta-actin: ACTB) were ob-
tained from Applied Biosystems (RETN: Hs00220767_m1, ACTB:
Hs99999903_01). Quantitation was performed using the DDCT
method with the median value of all samples as the control.

We also used paired t-tests (q  0.05; with Benjamini-Hochberg
FDR) to identify genes that showed significant differences between
samples at 0 and 12 weeks in the Chlorella group.
MetaCore v6.5 (Thomson Reuters; http://www.genego.com/)
was used to screen for diabetes-associated genes. These data were
2.8. Statistical analysis
Results for laboratory tests are represented as means ± S.E.Ms.
Unpaired t-tests (Welch's t-test) were used to compare the results
between the Chlorella and Placebo groups and paired t-tests were

Fig. 1. Flow chart of a randomized placebo-controlled study. One hundred thirty-five subjects who were diagnosed as borderline diabetes were recruited for this study. Sixty
subjects were randomly assigned Placebo and Chlorella group except for 75 subjects excluded on the basis of screening criteria. Of the 3 subjects who discontinued this study
because of voluntary. Of the 57 subjects (28 Chlorella, 29 placebo) completed treatment.
e98 H. Itakura et al. / Clinical Nutrition ESPEN 10 (2015) e95ee101

Fig. 2. Gene selection flowchart used to select six diabetes-associated genes. A) Inter-group comparisons: Comparisons of week 12 data with week 0 data (12e0 week) within the
Chlorella and Placebo groups. Welch's t-tests (P  0.05; without multiple testing corrections) resulted in 562 probes that showed changes between the Chlorella and Placebo groups.
B) Intra-group comparisons: Paired t-tests (P  0.05; with multiple testing corrections) were performed for the Chlorella and Placebo groups for those genes that exhibited
expression differences between weeks 0 and 12. These two data sets obtained with the Chlorella and Placebo groups resulted in 2181 probes that were uniquely related to Chlorella
ingestion. Finally, genes identified by Welch's tests (562) and t-tests (2181) were compared, of which 252 probes (222 genes) were in common. These genes were subsequently
compared with those listed in Metacore v6.5 (Type 2 diabetes-associated genes), which resulted in 6 diabetes-associated genes.

used for within-group comparisons with the week 0 control results.
P-value and q-value of 0.05 were considered significant. Pearson
correlation analysis was used to investigate associations between
gene expressions and laboratory test values. Correlation analysis
was also used to assess relationships between differences (amounts
of change) in resistin gene expression levels between weeks 12 and
0 that had been calculated from microarray data and the differences
(amounts of change) in laboratory test values between weeks 12
and 0.
3. Results
3.1. Subject characteristics
Table 1 shows the clinical characteristics of our study subjects,
except for two in the Chlorella group and one in the Placebo group
who withdrew their consent. There were no dropouts because of
adverse events caused by the study foods. Twenty-eight Chlorella-
group and 29 Placebo group subjects completed the 12-week
ingestion course (Fig. 1). There were no significant differences in
the clinical characteristics between the two groups, including
changes in body weight, body mass index (BMI) values, or
abdominal circumferences for those who completed the test.
3.2. Gene expression analyses for Chlorella dependent genes
Gene expression was determined as described in Methods. We
excluded some genes that showed significant differences in
expression between the Chlorella and Placebo groups at week
0 because these likely reflected individuals' characteristics. The
remaining 17,511 probe set were subjected to further analyses. First,
we identified those genes that showed changes between weeks 12
and 0 within the Chlorella and Placebo groups and subsequently
used unpaired Welch's t-tests (P  0.05). This showed that 562
probes had different expressions between these two groups
(Fig. 2A).

Table 2
Six diabetes-associated genes identified based on their significant expression changes induced by Chlorella ingestion.a

DChlorella 12e0 week vs. Chlorella Placebo

DPlacebo 12e0 week 12e0 week 12e0 week

Entrez GeneID Gene Symbol GeneName P valuea Fold change Regulation q valueb

56729 RETN Resistin 0.014 1.20 Down 0.041 0.73

5187 PER1 period homolog 1 (Drosophila) 0.019 1.19 Up 0.018 0.75
26053 AUTS2 autism susceptibility candidate 2 0.001 1.19 Up 0.015 0.51
7840 ALMS1 Alstrom syndrome 1 0.006 1.14 Up 0.017 0.70
3949 LDLR low density lipoprotein receptor 0.013 1.12 Up 0.003 0.96
5467 PPARD peroxisome proliferator-activated receptor delta 0.044 1.06 Up 0.021 0.75

Six genes were extracted as diabetes specific genes that reflected the effects of Chlorella by comparing 252 genes (Fig. 2).
Listed in descending order of differences in fold-changes between the Placebo and Chlorella groups.
a
P value from Fig. 2A: Welch's t-test.
b
q value from Fig. 2B: paired t-test.
 H. Itakura et al. / Clinical Nutrition ESPEN 10 (2015) e95ee101 e99

Fig. 3. Differences in resistin gene expression between those in the Chlorella and Placebo groups. A) Circles indicate each individual's resistin gene expression at weeks 0 and 12.
Paired t-tests were used for intra-group comparisons (weeks 0 and 12). Unpaired t-tests (Welch's t-test) were used to compare the D12e0 week data for the Chlorella and Placebo
groups. B) Plot of D12e0 week data along with their differences. Data for individuals were aligned in order of resistin expression (D12e0 week) values. The black and gray straight
lines represent the 75th percent levels of the Chlorella and Placebo group, respectively, and the broken lines represent the 25th percentile levels.

In parallel with this, we used paired t-tests (q  0.05) for those
genes that showed expression changes between weeks 0 and 12,
respectively, within the Chlorella and Placebo groups. With this we
obtained 366 probes whose expressions differed between weeks
0 and 12 in the Placebo group, and 2439 probes whose expressions
differed between weeks 0 and 12 in the Chlorella group. Chlorella
ingestion seemed to significantly affected gene expression in blood
cells. We subsequently used the 2181 probes with different ex-
pressions for further analyses (Fig. 2B).
The 562 probes obtained by Welch's t-tests and the 2181 probes
obtained by paired t-tests were compared, of which 252 probes
(222 genes) that were observed to be common between these two
tests were further examined for those that were related to diabetics
using MetaCore v6.5 (type 2 diabetes-associated genes). Table 2
shows six genes that were identified using this procedure.
By examining these six genes, more than two-thirds of those in
the Chlorella ingestion group exhibited resistin gene down-
regulation. It is known that resistin is secreted by adipocytes and is
a link between obesity and type 2 diabetes [15]. As will be discussed
later, the resistin gene is also active in macrophages [16].
Five other diabetes-associated genes were activated by Chlorella
ingestion. These genes included: the LDL receptor (LDLR) gene [17]
whose products enhance LDL intake from plasma; PER1, one of the
clock genes whose expression is known to be low in diabetics [18];
PPARD that regulates blood glucose and neutral fat concentrations
[19]; and ALMS1 whose product is known to control insulin pro-
duction in the pancreas [20].
We subsequently focused on alterations in resistin gene
expression. Fig. 3A shows that almost all of those in the Chlorella
ingestion group exhibited downregulated expression of this gene
(p ¼ 0.01). In contrast, nearly half of those in the Placebo ingestion
group exhibited this genes' expression in the opposite direction
(p ¼ 0.73). Welch's t-test results with 12e0 week data were 0.04
when we plotted the this data in descending order as shown in
Fig. 3B; downregulation of resistin gene expression in the Chlorella
ingestion group was more readily apparent.
To further confirm this we used qRT-PCR with 0- and 12-week
samples in the Chlorella group. The difference was significant
(p ¼ 0.01), whereas there was no significant difference in the Pla-
cebo group (p ¼ 0.90; data not shown). All of these results showed
that Chlorella ingestion suppressed resistin gene expression, which
reflected that Chlorella ingestion had a favorable affect by “calming
down” a diabetes predisposition, at least in part (see below).
Resistin levels are known to be correlated with pro-
inflammatory gene activities. Thus, we examined the blood levels
of IL-6 and TNF-a. These results are shown in Fig. 4 and demon-
strate that the levels of these inflammation-related cytokines were
low when the resistin gene expression level was also low. HbA1c,
which reflects the severity of diabetes, was also low when resistin
expression was low. In comparison, adiponectin receptor 1 (ADI-
POR1), whose gene expression should be inversely correlated with
resistin gene expression, was inversely expressed, as expected
(Fig. 5).
4. Discussion
We recruited middle-aged Japanese men who were borderline
diabetics for this randomized, placebo-controlled, double-blind
study. We compared the gene expressions in their blood cells at
0 and 12 weeks between the Chlorella and Placebo ingestion
groups. The effects of Chlorella ingestion were analyzed by Welch's
t-tests and paired t-tests. We subsequently focused on six genes
that are well correlated with type 2 diabetes.
Among these six genes, resistin gene expression in blood cells
was suppressed in those who had ingested Chlorella. This gene was
originally discovered in mice as an adipose tissue-derived hormone
that linked obesity to insulin resistance [15]. The blood levels of
resistin are known to be related to diabetes or insulin resistance in
mice and humans [15,21]. Recent studies demonstrated that human
resistin was produced not only by adipocytes but also by cells of the
immune system, particularly monocytes and macrophages [16,22].
The close correlation between resistin levels and inflammation [23]
e100 H. Itakura et al. / Clinical Nutrition ESPEN 10 (2015) e95ee101
Fig. 4. Correlations between differences in resistin gene expression level (Dresistin:
12e0 week) in the Chlorella group against differences in marker levels (D: 12e0 week).
(A) DHbA1c: 12 week HbA1c value minus 0 week HbA1c value, (B) DIL-6 (D: 12e0
week), and (C) DTNF-a (D: 12e0 week). Rhombuses indicate data for each individual.
strongly suggests that it is involved in chronic microinflammation,
which leads to obesity, metabolic syndrome, and diabetes [24]. In
addition, resistin has been shown to play a role in controlling hy-
pothalamic and peripheral lipid metabolism and in regulating food
intake [25]. Thus, resistin expression in immune cells may be a
means to monitor the progression of patients toward diabetes [26].
Cytokines that reflected inflammation, including TNF-a and IL-6,
were at lower levels in the blood, which paralleled the reduced
resistin gene expression (Fig. 4). Furthermore, HbA1c levels, a
marker of diabetic patient status, also paralleled these changes.
Adiponectin is an important hormone for improving insulin resis-
tance, primarily by promoting fatty acids and glucose catabolism
through AMPK activation in muscle and liver [27], should have
increased under our conditions. However, the blood levels of this
hormone did not show an increase for all of our study subjects. On
the other hand, Fig. 5 was shown which ADIPOR1 increased in
Fig. 5. Correlation between changes in resistin gene expression level (D12e0 week)
and adiponectin gene expression level (D12e0 week) in the Chlorella group. Rhom-
buses indicate different individuals.
parallel with reduced resistin gene expression, as expected.
ADIPOR1 gene levels have been reported to be inversely correlated
with insulin resistance [28]. All of these findings are in line with the
favorable effects of Chlorella for improvements in diabetes-
predisposed people.
Studies of this type are affected by fluctuations that reflect the
heterogeneity of human participants in conjunction with differ-
ences in individuals' physiological conditions. Thus, it was not
surprising to find that some showed “favorable physiological con-
ditions” without ingesting the test material, whereas others
showed “unfavorable physiological conditions” despite having
ingested the test material. Our data with resistin at week 0 could
have been affected by the genetic characteristics of our participants.
For example, it is known that 420G promoter recognition by Sp1/
3 is affected by increased serum resistin levels [29]. The G/G ge-
notype of the resistin gene reportedly induces diabetes mellitus
[30].
In addition, human physiological conditions are highly
individual-specific and far from uniform, in contrast to that with
experimental animals. For example, at week 0 when our test star-
ted, the resistin levels among our participants were anywhere be-
tween the highest and lowest values.
We did not find any significant improvement associated with
Chlorella ingestion in terms of HbA1c levels or with blood glucose
levels (measured under non-fasting conditions) at the end of 12
weeks. It was likely that studying mild pre-diabetic people for 12
weeks was not sufficient to achieve significant improvements.
Nevertheless, by focusing on blood cell resistin gene expression and
comparing data before and after our experimental intervention, we
detected some subtle physiological changes.
Statement of authorship
The authors' responsibilities were as follows-HI: study concep-
tion and design, data acquisition, drafting of the manuscript,
analysis and interpretation of the data, critical revision of the
manuscript for important intellectual content, and primary re-
sponsibility for final the content; MK: in the microarray experi-
ment, data acquisition, statistical analysis, and revision of the
manuscript for important intellectual content; and SN: in the
microarray experiment, the quantitative polymerase chain reaction
confirmation experiment, statistical analysis, study conception and
design, critical revision of the manuscript for important intellectual
content, and supervision. All authors read and approved the final
manuscript.
Conflict of interest statement
All authors have nothing to report.
Funding sources
This study was supported by Sun Chlorella Corp.
Acknowledgment
Study materials and financial support were provided by Sun
Chlorella Corp. This study was supported by Sun Chlorella Corp. and
registered at umin.ac.jp as UMIN000009130.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.clnesp.2015.04.002.
 H. Itakura et al. / Clinical Nutrition ESPEN 10 (2015) e95ee101 e101

References [17] Flock MR, Green MH, Kris-Etherton PM. Effects of adiposity on plasma lipid
response to reductions in dietary saturated fatty acids and cholesterol. Adv
Nutr 2011;2:261e74.
[1] Jacobson AM. Impact of improved glycemic control on quality of life in pa-
[18] Ando H, Takamura T, Matsuzawa-Nagata N, Shima KR, Eto T, Misu H, et al.
tients with diabetes. Endocrine practice. Off J Am Coll Endocrinol Am Assoc
Clock gene expression in peripheral leucocytes of patients with type 2 dia-
Clin Endocrinol 2004;10:502e8.
betes. Diabetologia 2009;52:329e35.
[2] de Groot M, Anderson R, Freedland KE, Clouse RE, Lustman PJ. Association of
[19] Brunmair B, Staniek K, Dorig J, Szocs Z, Stadlbauer K, Marian V, et al. Activa-
depression and diabetes complications: a meta-analysis. Psychosom Med
tion of PPAR-delta in isolated rat skeletal muscle switches fuel preference
2001;63:619e30.
from glucose to fatty acids. Diabetologia 2006;49:2713e22.
[3] Forte V, Pandey A, Abdelmessih R, Forte G, Whaley-Connell A, Sowers JR, et al.
[20] Minton JA, Owen KR, Ricketts CJ, Crabtree N, Shaikh G, Ehtisham S, et al.
Obesity, diabetes, the Cardiorenal syndrome, and risk for Cancer. Cardiorenal
Syndromic obesity and diabetes: changes in body composition with age and
Med 2012;2:143e62.
mutation analysis of ALMS1 in 12 United Kingdom kindreds with Alstrom
[4] Lee M, Saver JL, Hong KS, Song S, Chang KH, Ovbiagele B. Effect of pre-diabetes
syndrome. J Clin Endocrinol metabolism 2006;91:3110e6.
on future risk of stroke: meta-analysis. BMJ 2012;344:e3564.
[21] Tokuyama Y, Osawa H, Ishizuka T, Onuma H, Matsui K, Egashira T, et al. Serum
[5] Marwick TH, Hordern MD, Miller T, Chyun DA, Bertoni AG, Blumenthal RS,
resistin level is associated with insulin sensitivity in Japanese patients with
et al. Exercise training for type 2 diabetes mellitus: impact on cardiovascular
type 2 diabetes mellitus. Metab e Clin Exp 2007;56:693e8.
risk: a scientific statement from the American Heart Association. Circulation
[22] Curat CA, Wegner V, Sengenes C, Miranville A, Tonus C, Busse R, et al. Mac-
2009;119:3244e62.
rophages in human visceral adipose tissue: increased accumulation in obesity
[6] Nkondjock A, Receveur O. Fish-seafood consumption, obesity, and risk of type
and a source of resistin and visfatin. Diabetologia 2006;49:744e7.
2 diabetes: an ecological study. Diabetes Metab 2003;29:635e42.
[23] Bokarewa M, Nagaev I, Dahlberg L, Smith U, Tarkowski A. Resistin, an adi-
[7] Mizoguchi TAY, Fujishima M. Effect of chlorella on the metabolic syndrome
pokine with potent proinflammatory properties. J Immunol 2005;174:
using Spontaneous obese mice. Med Biol 2009;153:327e33.
5789e95.
[8] Mizoguchi T, Takehara I, Masuzawa T, Saito T, Naoki Y. Nutrigenomic studies
[24] Festa A, D'Agostino Jr R, Tracy RP, Haffner SM. Elevated levels of acute-phase
of effects of Chlorella on subjects with high-risk factors for lifestyle-related
proteins and plasminogen activator inhibitor-1 predict the development of
disease. J Med Food 2008;11:395e404.
type 2 diabetes: the insulin resistance atherosclerosis study. Diabetes
[9] Senthilkumar T, Sangeetha N, Ashokkumar N. Antihyperglycemic, anti-
2002;51:1131e7.
hyperlipidemic and renoprotective effects of Chlorella pyrenoidosa in diabetic
[25] Nogueiras R, Novelle MG, Vazquez MJ, Lopez M, Dieguez C. Resistin: regula-
rats exposed to cadmium. Toxicol Mech Methods 2012.
tion of food intake, glucose homeostasis and lipid metabolism. Endocr Dev
[10] Nakamura S, Kobayashi M, Sugino T, Kajimoto O, Matoba R, Matsubara K.
2010;17:175e84.
Effect of exercise on gene expression profile in unfractionated peripheral
[26] Kaser S, Kaser A, Sandhofer A, Ebenbichler CF, Tilg H, Patsch JR. Resistin
blood leukocytes. Biochem Biophys Res Commun 2010;391:846e51.
messenger-RNA expression is increased by proinflammatory cytokines
[11] Ghosh S, Dent R, Harper ME, Gorman SA, Stuart JS, McPherson R. Gene
in vitro. Biochem Biophys Res Commun 2003;309:286e90.
expression profiling in whole blood identifies distinct biological pathways
[27] Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, Uchida S, et al. Adiponectin
associated with obesity. BMC Med genomics 2010;3:56.
stimulates glucose utilization and fatty-acid oxidation by activating AMP-
[12] Takamura T, Honda M, Sakai Y, Ando H, Shimizu A, Ota T, et al. Gene expression
activated protein kinase. Nat Med 2002;8:1288e95.
profiles in peripheral blood mononuclear cells reflect the pathophysiology of
[28] Tsuchida A, Yamauchi T, Ito Y, Hada Y, Maki T, Takekawa S, et al. Insulin/Foxo1
type 2 diabetes. Biochem Biophys Res Commun 2007;361:379e84.
pathway regulates expression levels of adiponectin receptors and adiponectin
[13] Boomgaarden I, Egert S, Rimbach G, Wolffram S, Muller MJ, Doring F. Quer-
sensitivity. J Biol Chem 2004;279:30817e22.
cetin supplementation and its effect on human monocyte gene expression
[29] Osawa H, Yamada K, Onuma H, Murakami A, Ochi M, Kawata H, et al. The G/G
profiles in vivo. Br J Nutr 2010;104:336e45.
genotype of a resistin single-nucleotide polymorphism at -420 increases type
[14] Schmidt S, Willers J, Stahl F, Mutz KO, Scheper T, Hahn A, et al. Regulation of lipid
2 diabetes mellitus susceptibility by inducing promoter activity through
metabolism-related gene expression in whole blood cells of normo- and dys-
specific binding of Sp1/3. Am J Hum Genet 2004;75:678e86.
lipidemic men after fish oil supplementation. Lipids health Dis 2012;11:172.
[30] Osawa H, Onuma H, Ochi M, Murakami A, Yamauchi J, Takasuka T, et al.
[15] Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, et al. The
Resistin SNP-420 determines its monocyte mRNA and serum levels
hormone resistin links obesity to diabetes. Nature 2001;409:307e12.
inducing type 2 diabetes. Biochem Biophys Res Commun 2005;335:
[16] Patel L, Buckels AC, Kinghorn IJ, Murdock PR, Holbrook JD, Plumpton C, et al.
596e602.
Resistin is expressed in human macrophages and directly regulated by PPAR
gamma activators. Biochem biophysical Res Commun 2003;300:472e6.