Plasma free amino acid profiles evaluate risk of metabolic syndrome, diabetes, dyslipidemia, and hypertension in a large Asian population
Yamaguchi et al. Environmental Health and Preventive Medicine
Plasma free amino acid profiles evaluate risk of metabolic syndrome, diabetes, dyslipidemia, and hypertension in a large Asian population
Natsu Yamaguchi 0
MH Mahbub 0
Hidekazu Takahashi 0
Ryosuke Hase 0
Yasutaka Ishimaru 0
Hiroshi Sunagawa 0
Mikiko Kobayashi- Miura
Tsuyoshi Tanabe 0
0 Department of Public Health and Preventive Medicine, Graduate School of Medicine, Yamaguchi University , 1-1-1 Minami-Kogushi, Ube, Yamaguchi 755-8505 , Japan
Background: Recently, the association of plasma free amino acid (PFAA) profile and lifestyle-related diseases has been reported. However, few studies have been reported in large Asian populations, about the usefulness of PFAAs for evaluating disease risks. We examined the ability of PFAA profiles to evaluate lifestyle-related diseases in so far the largest Asian population. Methods: We examined plasma concentrations of 19 amino acids in 8589 Japanese subjects, and determined the association with variables associated with obesity, blood glucose, lipid, and blood pressure. We also evaluated the PFAA indexes that reflect visceral fat obesity and insulin resistance. The contribution of single PFAA level and relevant PFAA indexes was also examined in the risk assessment of lifestyle-related diseases. Results: Of the 19 amino acids, branched-chain amino acids and aromatic amino acids showed association with obesity and lipid variables. The PFAA index related to visceral fat obesity showed relatively higher correlation with variables than that of any PFAA. In the evaluation of lifestyle-related disease risks, the odds ratios of the PFAA index related to visceral fat obesity or insulin resistance with the diseases were higher than most of those of individual amino acid levels even after adjusting for potential confounding factors. The association pattern of the indexes and PFAA with each lifestyle-related disease was distinct. Conclusions: We confirmed the usefulness of PFAA profiles and indexes as markers for evaluating the risks of lifestyle-related diseases, including diabetes mellitus, metabolic syndrome, dyslipidemia, and hypertension in a large Asian population.
Amino acids; Lifestyle-related diseases; Metabolic syndrome; Diabetes mellitus; Dyslipidemia; Hypertension
Lifestyle-related diseases are pathophysiological states
that include metabolic syndrome, diabetes mellitus
(DM), dyslipidemia, hypertension and gout, and often
lead to cardiovascular disease [1, 2]. Excess visceral fat
and hyperinsulinemia are regarded as risk factors of
lifestyle-related diseases [3, 4], and early detection of
such factors is important for prevention of the diseases.
Because of a shift towards the Western lifestyle and diet,
the prevalence of lifestyle-related diseases has
significantly increased in Asian populations . For early
detection of lifestyle-related diseases, the development of
useful biomarkers is essential which would also help in
better understanding of the disease pathophysiology and
early intervention to avoid the progression of diseases
and deterioration of patients? conditions .
Recent reports have shown that plasma free amino
acid (PFAA) profile can serve as an effective biomarker
? The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
for detection of lifestyle-related diseases [6?9]. PFAA
profile was also reported to be useful for identifying
diabetic patients at risk of developing cardiovascular
diseases [10, 11]. Although these reports suggested that
changes of amino acid metabolism play important roles
in the pathogenesis of lifestyle-related diseases, the
number of subjects in those studies was limited.
Furthermore, we previously reported the construction of PFAA
index 1 and index 2, which evaluated visceral fat area
(VFA) and insulin resistance, respectively, and revealed
their association with variables of lifestyle-related
diseases and usefulness of such indexes for predicting the
future risk of developing the diseases including
metabolic syndrome, DM, dyslipidemia, and hypertension
. However, it is necessary to confirm the capabilities
of the PFAA indexes to evaluate the risk of
lifestylerelated diseases in a large Japanese population.
Therefore, the purpose of this study was to examine
the ability of PFAA and relevant indexes for evaluating
lifestyle-related diseases in so far the largest sample of
an Asian population for such type of studies.
The study was conducted in accordance with the
Declaration of Helsinki, and the institutional review
board of Shimane University (20100129-3) and Yamaguchi
University (H25-26-2) approved the current study
protocol. A total of 8589 Japanese subjects who underwent the
health examination during 2009 to 2011 in Shimane
prefecture and gave informed consent to participate in the
study, were included.
Measurement of metabolic variables and quantification of
Blood samples were taken from the subjects after an 8-h
fast. Serum levels of total cholesterol (T-CHO),
highdensity lipoprotein cholesterol (HDL-C), low- density
lipoprotein cholesterol (LDL-C), and triglyceride (TG)
were determined enzymatically. Free plasma glucose
(FPG) was measured with the hexokinase method, and
HbA1c was determined using the latex agglutination
immunoassay. Plasma amino acid concentrations were
analyzed following the protocol previously described
elsewhere [12?16]. Briefly, blood samples (5 mL) were
collected from forearm veins after overnight fasting into
tubes containing disodium
ethylenediaminetetraacetate that were immediately placed on ice. The
plasma was prepared by centrifugation at 3000 r.p.m.
at 4 ?C for 15 min and then stored at ?80 ?C until
analysis. The plasma amino acid concentrations were
measured by high-performance liquid chromatography?
electrospray ionization mass spectrometry followed by
precolumn derivatization as previously described [13?16].
The following 19 amino acids were measured: Alanine
(Ala), Arginine (Arg), Asparagine (Asn), Citrulline (Cit),
Glutamine (Gln), Glycine (Gly), Histidine (His), Isoleucine
(Ile), Leucine (Leu), Lysine (Lys), Methionine (Met),
Ornithine (Orn), Phenylalanine (Phe), Proline (Pro),
Serine (Ser), Threonine (Thr), Tryptophane (Trp),
Tyrosine (Tyr), and Valine (Val). We did not perform the
measurements of other genetically-encoded amino acids like
glutamate, aspartate, and cysteine due to their instability
in the blood .
In this study, metabolic syndrome was defined according
to the following Japanese diagnostic criteria for the
syndrome: visceral obesity (waist ? 85 cm in males and
?90 cm in females) plus at least 2 of the following three
components: (1) HDL-C < 40 mg/dL, TG ? 150 mg/dL,
or the use of medication for dyslipidemia; (2) FPG ?
110 mg/dL or the use of medication for DM; and (3)
blood pressure ?130/85 mmHg or the use of
antihypertensive medication. DM was defined in patients with
FPG ?126 mg/dL, HbA1c ? 6.5%, or those who were
taking medication for DM. Dyslipidemia was defined in
individuals with fasting LDL-C ? 140 mg/dL, HDL-C <
40 mg/dL, TG ? 150 mg/dL, or those who were taking
medication for dyslipidemia. Hypertension was defined
in patients with systolic blood pressure (SBP) ?140 or
diastolic blood pressure (DBP) ? 90 mmHg or those who
were taking antihypertensive medications.
Calculation of PFAA indexes
In this study, we used the amino acid index 1 and index
2 constructed and validated in a previous study .
The amino acid index 1 is the multiple linear regression
model with variable selection to model the relationships
between the PFAA profiles with the visceral fat area,
consisting of Leu, Ala, Tyr, Asn, Trp, and Gly. The
amino acid index 2 is the multiple linear regression
model with variable selection to model the relationships
between the PFAA profiles with 2-h post-challenge
insulin levels (Ins120 min), consisting of Ile, Ala, Tyr, Phe,
Met, and His. Therefore, each of these PFAA indexes
(index 1 and index 2) is a single dimension that contains
information on multidimensional PFAA profiles. Such
compression of information on PFAA profiles allows
maximization of the discrimination between patients
and control subjects.
Correlation between PFAAs and metabolic variables
Correlation analysis between each of single PFAA
concentration and PFAA indexes and metabolic variables
was performed as previously described  by using the
Pearson product-moment correlation coefficient. In
addition, two-dimensional hierarchical cluster analysis
that was based on the correlation coefficient matrix
between the PFAA concentrations and additional measured
variables was performed. For each single PFAA and the
PFAA indexes, three different models were used with or
without adjustments for variables as follows: model 1)
without adjusting, model 2) adjusted for age and
gender, model 3) adjusted for age, gender, and body mass
Association between PFAAs and lifestyle related diseases
We examined the relationships of the PFAA profiles to
lifestyle-related diseases to determine whether each
PFAA index and single PFAA concentration were related
to DM, metabolic syndrome, dyslipidemia, and
hypertension. The PFAA indexes and all of the amino acids
were scaled to multiples of 1 SD. A logistic regression
analysis was used to assess the contribution of each
PFAA index and single PFAA concentration as
continuous variables in the evaluation of these diseases. The
logistic regression analysis was performed with
adjustment for age and gender. Furthermore, to exclude the
cross-over effects among diseases on the single PFAA
level and each PFAA index, further adjustments were
performed as follows: metabolic syndrome for age,
gender and BMI; DM for age, gender, BMI, LDL-C, HDL-C,
TG, SBP and DBP; dyslipidemia for age, gender, BMI,
FPG, HbA1c, SBP and DBP; hypertension for age,
gender, BMI, FPG, HbA1c, LDL-C, HDL-C and TG.
A two-sided probability value of p < 0.01 was
considered to be statistically significant. R version 3.1.3 [R Core
Team (2015). R: A language and environment for
statistical computing. R Foundation for Statistical Computing,
Vienna, Austria] was used for the statistical analyses.
All of the data were analyzed anonymously
throughout the study.
The numbers of subjects and their demographic and
clinical characteristics have been presented in Table 1.
Significant differences (p < 0.001 to 0.01) between the
diseased and non-diseased populations were observed
for all variables except for LDL cholesterol for DM and
Amino acid concentrations in the study populations
are shown in Table 2. Concentrations of PFAA differed
significantly between the diseased and non-diseased
subjects for metabolic syndrome, DM, dyslipidemia, and
hypertension (p < 0.001 to 0.05). Overall, the
concentrations of PFAA were higher in the diseased compared to
the non-diseased; however, the concentrations of Gly
and Ser for all diseases (except Ser for DM) and Asn for
dyslipidemia showed the opposite trend.
We first examined the relationship between variables
that were associated with lifestyle-related diseases and
amino acids measured in this study (Table 3). The
variables are obesity-related (BMI, waist circumference),
glucose-related (FPG, HbA1c), lipid-related (HDL-C,
LDL-C, and TG), and blood pressure-related (SBP, DBP).
Branched-chain amino acids (BCAA; Val, Leu, Ile)
showed moderately positive correlation with both
obesity variables and with TG. All other correlations of
BCAA with the rest variables were positive but weak.
Like BCAA, aromatic amino acids (AAA; Tyr, Phe, Trp)
also showed similar trends of correlations: positive
(moderate to weak) correlation with both obesity
variables and with TG, and weakly positive correlations with
others. Among gluconeogenic amino acids (Ala, Gly,
Gln, Ser), Ala showed moderately positive correlation
with several variables including BMI, waist circumstance,
blood glucose, and TG. Conversely, the correlation of
Gly, Gln, and Ser was very weak or negative with
almost all variables. Correlations of other single amino
acids with lifestyle-related diseases were either very
low or did not show any specific trend. Both PFAA
indexes exhibited moderate to weak positive
correlations with most of the variables associated with
lifestyle-related diseases. Overall, the correlations were
higher for the PFAA indexes compared with those for
the single amino acids.
We then examined the capability of PFAA to evaluate
the risk of metabolic syndrome, DM, dyslipidemia, and
hypertension after adjustments for age and gender
(Table 4). All the BCAA showed significant positive
associations with all four diseases examined [OR between
1.23 and 2.16; 95% CI between 1.15 and 1.98 (lower) and
1.30 to 2.36 (upper); p < 0.001]. Especially, the odds
ratios of BCAA with metabolic syndrome, DM and
dyslipidemia were high, at least 1.7. Also, the odds ratios of
AAA showed significant positive association with all four
diseases [OR between 1.11 and 1.65; 95% CI between
1.05 and 1.52 (lower) and 1.17 to 1.78 (upper); p < 0.001]
except Trp in DM. The odds ratios shown by AAA were
relatively low compared with those by BCAA.
Gluconeogenesis-related amino acids showed a typical
pattern of odds ratio for all the four diseases, positive
association for Ala [OR between 1.37 and 2.15; 95% CI
between 1.30 and 1.52 (lower) and 1.98 to 2.33 (upper);
p < 0.001] and inverse association for Gly, Gln, and Ser
[OR between 0.57 and 0.85; 95% CI between 0.51 and
0.80 (lower) and 0.64 to 0.89 (upper); p < 0.001] except
for Ser in DM and Gln in dyslipidemia (Table 3). Then
we confirmed the capabilities of the PFAA index 1 [OR
between 1.52 and 3.23; 95% CI between 1.44 and 2.92
(lower) and 1.61 to 3.57 (upper); p < 0.001] and index 2
[OR between 1.48 and 2.37; 95% CI between 1.39 and
2.16 (lower) and 1.57 to 2.61 (upper); p < 0.001] to
evaluate the significant positive association with
lifestylerelated diseases (Table 4). The index 1 showed higher
Table 3 Correlation between PFAA profiles and metabolic variables
odds ratio than that of any amino acid with metabolic
syndrome, DM, and hypertension. The index 2 showed
higher odds ratio than that of each amino acid with
metabolic syndrome and hypertension.
Next we confirmed the capabilities of the PFAA to
evaluate their association with lifestyle-related
diseases after excluding the probable cross-over effects
among diseases on the single PFAA level and each
PFAA index (Table 5). For this purpose, the diseases
were adjusted for different relevant factors for each
disease separately. After such adjustments, the
typical pattern of association for BCAA persisted with
all diseases except hypertension. In contrast, after
adjustments, the odds ratios of AAA lost significant
positive associations for all diseases except Trp in
dyslipidemia. On the other hand, most of the
gluconeogenesis-related amino acids retained their
typical pattern of association with the diseases. Also,
positive associations of both PFAA indexes remained
statistically significant with all lifestyle-related
diseases (Table 5).
In this study, we have examined the relationship
between PFAA profiles and lifestyle-related disease risks
with more than 8000 subjects, so far the largest Asian
population for such type of studies. We confirmed that
PFAA profiles and the indexes generated from them are
useful for evaluating lifestyle-related diseases. Our
analysis showed that concentrations of several single PFAA
and PFAA indexes could identify the subjects who are at
risks for metabolic syndrome, DM, dyslipidemia, and
Recently, an increased number of reports on the
association between PFAA profiles and lifestyle-related
diseases are being observed [11?13]. For example, BCAAs
especially Val and Leu, have been proposed as a
cardiometabolic risk marker independent of BMI category
. Furthermore, BCAAs have been linked to the
metabolic syndrome, insulin resistance or type 2 DM [12, 17].
Interactions of excess BCAA and lipids may lead to the
development of ?-cell dysfunction, which drives the
transition from the obese, insulin-resistant state to type
Values are odds ratios (95% confidence intervals) as a continuous variable per SD increment for developing metabolic syndrome, DM, dyslipidemia, or hypertension
from logistic regressions
2 DM . With respect to such factors like insulin sen
sitivity and visceral obesity, there are differences between
Asian and Western populations [18, 19]. But the number
of studies focusing on these issues is limited [12, 20].
As revealed in this study, there was a consistent
increase in the concentrations of BCAA in all four
lifestyle-related diseases. Also, BCAA exhibited
significant positive association with all those diseases. Such an
observation is consistent with the findings of other
research works [12, 13]. As the underlying mechanisms, it
has been suggested that a decrease in insulin activity
and utilization of amino acids in muscles induces
reduced uptake of BCAAs into muscles resulting in
increased BCAA levels in lifestyle-related diseases .
Newgard et al. postulated that the rise in circulating
BCAA in obese and insulin-resistant subjects is partially
caused by a decline in their catabolism in adipose tissue
possibly via down-regulation of the BCAA catabolic
enzymes through the suppression of peroxisome
proliferator-activated receptor-? (PPAR-?) signaling in
such metabolic adaptation . All these might have
been reflected in our findings of significant positive
association of BCAA with lifestyle-related diseases. However,
although metabolic syndrome, DM, dyslipidemia, and
hypertension showed positive association with BCAA in
this study after adjusting for age and gender,
hypertension lost the association after further adjustments for the
disease-specific factors. These results might suggest that
in addition to the currently known risk factors like diet
or nutrition, differing metabolism for different amino
acids probably contributes to increased risk of
The underlying mechanisms for the observed elevation
in AAA concentrations remain unclear. However,
repression of tyrosine aminotransferase during states of insulin
resistance and DM may result in the increased levels of
circulating Tyr and Phe as observed in our study [12, 22].
In our analysis, metabolic syndrome, dyslipidemia, and
hypertension showed similar patterns for association with
gluconeogenesis-related amino acids. The diseases
positively correlated with Ala, and negatively with Gly, Gln
and Ser except Ser in DM and Gln in dyslipidemia. It was
reported that under condition of hyperinsulinemia,
increased protein turnover causes greater supply of
gluconeogenic amino acids to the liver, leading to the
reductions of Gly, Gln and Ser . Additionally,
gluconeogenesis precursors including Ala have been reported to
rise in those with deteriorating glucose tolerance .
The present study has confirmed that PFAA indexes
which reflect VFA and insulin resistance, respectively
, could assess lifestyle-related diseases investigated in
the largest Japanese population. In this study, index 1
was associated with most of the variables more positively
than any single amino acid, and the tendency was more
apparent than that of index 2. The association pattern of
the indexes and PFAA with each lifestyle-related disease
is distinct. Although hypertension showed no association
with BCAA and AAA after adjustments for exclusion of
cross-over effects among diseases, corresponding odds
ratios for both index 1 and index 2 were significant. Of
the diseases examined, only dyslipidemia showed higher
odds ratios with BCAA than with index 1. These results
revealed the distinct contribution of amino acid
metabolism to the risk of each lifestyle-related disease. Further
analysis of the relationship between the amino acid
metabolisms and disease risk may lead to the understanding
of pathophysiology, diagnosis and prevention of
Our analysis showed moderate correlation between
PFAA indexes and hypertension, which was previously
shown to be weak . The difference between the
results might be due to a higher number of subjects in this
study. However, it would be interesting to know the
effects of salt intake on the relationship between
hypertension and PFAA profiles in future studies.
The findings of our study should be interpreted in
light of the limitation that we could not include the
information related to the lifestyle factors such as physical
activity and diet due to limited available data on these,
which may have affected the PFAA profiles observed in
Our study confirmed the usefulness of PFAA profiles as
markers for evaluating the risks of lifestyle-related
diseases, including DM, metabolic syndrome, dyslipidemia,
and hypertension in a large Asian population.
AAA: Aromatic amino acids; Ala: Alanine; Arg: Arginine; Asn: Asparagine;
BCAA: Branched-chain amino acids; BMI: Body mass index; Cit: Citrulline;
DBP: Diastolic blood pressure; DM: Diabetes mellitus; FPG: Free plasma glucose;
Gln: Glutamine; Gly: Glycine; HbA1c: Haemoglobin A1c; HDL-C: High-density
lipoprotein cholesterol; His: Histidine; Ile: Isoleucine; LDL-C: Low-density
lipoprotein cholesterol; Leu: Leucine; Lys: Lysine; Met: Methionine;
Orn: Ornithine; PFAA: Plasma free amino acid; Phe: Phenylalanine; Pro: Proline;
SBP: Systolic blood pressure; Ser: Serine; T-CHO: Cholesterol;
TG: Triglyceride; Thr: Threonine; Trp: Tryptophane; Tyr: Tyrosine; Val: Valine;
VFA: Visceral fat area
Availability of data and materials
Please contact the corresponding author for data requests.
NY, YF, HA, MKM, HY and TT designed the study. NY, MM, HT, RH, YI, HS, HA,
MKM, HK, YF, HY and TT were involved in obtaining and surveying data. NY,
MM, HT, YF, HY, MY, SK, AI, MT, NK, MN and TT analyzed the data and performed
statistical analysis. All authors read and approved the final manuscript.
HY, MY, SK, AI, MT, NK, and MN are employees of Ajinomoto Co., Inc. TT, HA,
and YF received research grants from Ajinomoto Co., Inc. This does not alter
the authors? adherences to all of the journal policies. No other potential
conflicts of interest in relation to this article are declared.
Consent for publication
Ethics approval and consent to participate
The study was conducted in accordance with the Declaration of Helsinki,
and the institutional review board of Shimane University (20100129-3) and
Yamaguchi University (H25-26-2) approved the current study protocol. A
total of 8589 Japanese subjects who underwent the health examination
during 2009 to 2011 in Shimane prefecture and gave informed consent to
participate in the study, were included.
1. Tenenbaum A , Motro M , Schwammenthal E , Fisman E. Macrovascular complications of metabolic syndrome: an early intervention is imperative . Int J Cardiol . 2004 ; 97 : 167 - 72 .
2. Ma RC , Chan JC. Type 2 diabetes in East Asians: similarities and differences with populations in Europe and the United States . Ann N Y Acad Sci . 2013 ; 1281 : 64 - 91 .
3. Despres JP , Lemieux I. Abdominal obesity and metabolic syndrome . Nature . 2006 ; 444 : 881 - 7 .
4. Van Gaal LF , Mertens IL , De Block CE . Mechanisms linking obesity with cardiovascular disease . Nature . 2006 ; 444 : 875 - 80 .
5. Imaizumi A , Nishikawa N , Yoshida H , Yoneda J , Takahena S , Takahashi M , et al. Clinical implementation of metabolomics . In: Roessner U, editor. Metabolomics. Rijeka: InTech; 2012 . p. 289 - 314 .
6. McCormack SE , Shaham O , McCarthy MA , Deik AA , Wang TJ , Gerszten RE , Clish CB , Mootha VK , Grinspoon SK , Fleischman A. Circulating branchedchain amino acid concentrations are associated with obesity and future insulin resistance in children and adolescents . Pediatr Obes . 2013 ; 8 : 52 - 61 .
7. W?rtz P , M?kinen VP , Soininen P , Kangas A , Tukiainen T , Kettunen J , et al. Metabolic signatures of insulin resistance in 7,098 young adults . Diabetes . 2012 ; 61 : 1372 - 80 .
8. W?rtz P , Soininen P , Kangas A , R?nnemaa T , Lehtim?ki T , K?h?nen M , et al. Branched-chain and aromatic amino acids are predictors of insulin resistance in young adults . Diabetes Care . 2013 ; 36 : 648 - 55 .
9. Wang T , Larson M , Vasan R , Cheng S , Rhee E , McCabe E , et al. Metabolite profiles and the risk of developing diabetes . Nat Med . 2011 ; 17 : 448 - 53 .
10. Kume S , Araki S , Ono N , Shinhara A , Muramatsu T , Araki H , et al. Predictive properties of plasma amino acid profile for cardiovascular disease in patients with type 2 diabetes . PLoS One . 2014 ; 9 ( 6 ): e101219 .
11. Mangge H , Zelzer S , Pr?ller F , Schnedl WJ , Weghuber D , Enko D , et al. Branched-chain amino acids are associated with cardiometabolic risk profiles found already in lean, overweight and obese young . J Nutr Biochem . 2016 ; 32 : 123 - 7 .
12. Yamakado M , Nagao K , Imaizumi A , Tani M , Toda A , Tanaka T , et al. Plasma free amino acid profiles predict four-year risk of developing diabetes, metabolic syndrome, dyslipidemia, and hypertension in Japanese population . Sci Rep . 2015 ; 5 : 11918 .
13. Yamakado M , Tanaka T , Nagao K , Ishizaka Y , Mitushima T , Tani M , et al. Plasma amino acid profile is associated with visceral fat accumulation in obese Japanese subjects . Clin Obes . 2012 ; 2 : 29 - 40 .
14. Miyagi Y , Higashiyama M , Gochi A , Akaike M , Ishikawa T , Miura T , et al. Plasma free amino acid profiling of five types of cancer patients and its application for early detection . PLoS One . 2011 ; 6 ( 9 ): e24143 .
15. Shimbo K , Oonuki T , Yahashi A , Hirayama K , Miyano H. Precolumn derivatization reagents for high-speed analysis of amines and amino acids in biological fluid using liquid chromatography /electrospray ionization tandem mass spectrometry . Rapid Commun Mass Spectrom . 2009 ; 23 : 1483 - 92 .
16. Shimbo K , Yahashi A , Hirayama K , Nakazawa M , Miyano H. Multifunctional and highly sensitive precolumn reagents for amino acids in liquid chromatography/tandem mass spectrometry . Anal Chem . 2009 ; 81 : 5172 - 9 .
17. Newgard CB . Interplay between lipids and branched-chain amino acids in development of insulin resistance . Cell Metab . 2012 ; 15 : 606 - 14 .
18. Chiu KC , Chuang LM , Yoon C. Comparison of measured and estimated indices of insulin sensitivity and beta cell function: impact of ethnicity on insulin sensitivity and beta cell function in glucose-tolerant and normotensive subjects . J Clin Endocrinol Metab . 2001 ; 86 : 1620 - 5 .
19. Lim U , Ernst T , Buchthal SD , Latch M , Albright CL , Wilkens LR , Kolonel LN , Murphy SP , Chang L , Novotny R , Le Marchand L. Asian women have greater abdominal and visceral adiposity than Caucasian women with similar body mass index . Nutr Diabetes . 2011 ; 1 : e6 .
20. Takashina C , Tsujino I , Watanabe T , Sakaue S , Ikeda D , Yamada A , et al. Associations among the plasma amino acid profile, obesity, and glucose metabolism in Japanese adults with normal glucose tolerance . Nutr Metab (Lond) . 2016 ; 13 :5.
21. Pozefsky T , Felig P , Tobin JD , Soeldner JS , Cahill Jr GF . Amino acid balance across tissues of the forearm in postabsorptive man. Effects of insulin at two dose levels . J Clin Invest . 1969 ; 48 : 2273 - 82 .
22. Adams SH . Emerging perspectives on essential amino acid metabolism in obesity and the insulin-resistant state . Adv Nutr . 2011 ; 2 : 445 - 56 .
23. Chevalier S , Burgess SC , Malloy CR , Gougeon R , Marliss EB , Morais JA . The greater contribution of gluconeogenesis to glucose production in obesity is related to increased whole-body protein catabolism . Diabetes . 2006 ; 55 : 675 - 81 .
24. W?rtz P , Tiainen M , M?kinen VP , Kangas AJ , Soininen P , Saltevo J , et al. Circulating metabolite predictors of glycemia in middle-aged men and women . Diabetes Care . 2012 ; 35 : 1749 - 56 .