Metabolomics-Based Prospective Studies and Prediction of Type 2 Diabetes Mellitus Risks

Abstract

The preceding decade has witnessed an intense upsurge in the diabetic population across the world making type 2 diabetes mellitus (T2DM) more of an epidemic than a lifestyle disease. Metabolic disorders are often latent for a while before becoming clinically evident, thus reinforcing the pursuit of early biomarkers of metabolic alterations. A prospective study along with metabolic profiling is the most appropriate way to detect the early pathophysiological changes in metabolic diseases such as T2DM. The aim of this review was to summarize the different potential biomarkers of T2DM identified in prospective studies, which used tools of metabolomics. The review also demonstrates on how metabolomic profiling-based prospective studies can be used to address a concern like population-specific disease mechanism. We performed a literature search on metabolomics-based prospective studies on T2DM using the key words “metabolomics,” “Type 2 diabetes,” “diabetes mellitus”, “metabolite profiling,” “prospective study,” “metabolism,” and “biomarker.” Additional articles that were obtained from the reference lists of the articles obtained using the above key words were also examined. Articles on dietary intake, type 1 diabetes mellitus, and gestational diabetes were excluded. The review revealed that many studies showed a direct association of branched-chain amino acids and an inverse association of glycine with T2DM. Majority of the prospective studies conducted were targeted metabolomics-based, with Caucasians as their study cohort. The whole disease risk in populations, including Asians, could therefore not be identified. This review proposes the utility of prospective studies in conjunction with metabolomics platform to unravel the altered metabolic pathways that contribute to the risk of T2DM.

Introduction

Identification of different molecular targets of therapeutic significance has always been a priority in biomedical research. The quest to discover different molecules, such as RNAs, protein, metabolites, and so on, which can serve as biomarkers for predicting the progression of disease or the response to treatment, has significantly intensified. Technical advances in the field of transcriptomics, proteomics, and recently metabolomics have helped in streamlining the development of novel diagnostic markers.^1,2

Metabolomics can be explained as the study of metabolites, which are by-products of metabolism in biological samples.^1,3 Tools of metabolomics provide a direct functional readout of the physiological state of an organism, which was difficult to obtain using other omics approaches.^2,4 In a simple analogy, the genome can be referred to as a recipe, while proteome describes the ingredients required and metabolome would be the end product. Therefore, the metabolic profile acquired is the complete portrayal of the organism's phenotype.⁵ The tools used in metabolomic profiling include infrared and Raman spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, and a wide range of mass spectrometry (MS) based techniques. These tools help in obtaining a metabolic profile, which is dependent on the cellular and external environmental conditions which in turn may be altered under diseased conditions like diabetes mellitus (DM).^4,6

DM is a chronic metabolic disorder characterized by elevated blood glucose level, which can be due to insufficient insulin or inefficient insulin action. It is broadly classified into three—type 1, type 2, and gestational diabetes. Type 2 DM (T2DM) is the most prevalent form of diabetes characterized by insulin resistance and elevated glucose production.⁷ The metabolic makeup is altered due to persistent hyperglycemia that eventually leads to microvascular and macrovascular complications. It is now established that the prime cause of T2DM is associated with different lifestyle factors such as diet, lack of physical activity, obesity, and stress.⁸ T2DM patients, as well as individuals who are at risk, are advised to make several lifestyle changes, for instance, consumption of healthy diet, regular exercises, maintaining a normal body mass index (BMI), cessation of smoking, and so on, to prevent complications associated with diabetes, as well as to delay the onset of diabetes.⁸

Different facets, such as pathogenesis, diagnosis, complications, and so on, have been scrutinized persistently for centuries in diabetes research.⁹ While diagnosis and treatment of diabetes have been meticulously explored, studies on the discovery of early biomarkers of metabolic alterations involved in T2DM are only a handful. Devoid of timely identification and appropriate management, persons diagnosed with diabetes end up with one or more complications thereby accentuating the significance of early diagnosis of the disease.¹⁰ Tools of metabolomics will allow detection of pathophysiological alterations in body fluids, cells, or tissues and therefore could be a useful tool for disease diagnosis or biomarker detection. Many studies using the tools of metabolomics have reported different metabolic changes by comparing individuals with T2DM and without T2DM,^6,8,10,11 as well as obese and lean individuals.^12

–17 Among these, a prospective study along with metabolic profiling is a more pertinent way to detect early pathophysiological changes in T2DM.¹⁸

A prospective study involves following up a cohort of subjects for a specific time period. Metabolic diseases are often latent for a while before becoming clinically evident.¹⁸ For example, a considerable amount of pancreatic beta cells would be deficient by the time diagnosis of diabetes is made.¹⁹ Similarly, it is likely that the metabolic pathways are altered beforehand of a prediabetic state. Therefore it is necessary to conduct a prospective study along with metabolic profiling to monitor these changes. This review focuses on the prospective studies conducted hitherto in T2DM using tools of metabolomics and the subsequent avenues that can be pursued in this regard.

Furthermore, Asian Indians were shown to be more susceptible to T2DM compared to other populations. A healthy nondiabetic Indian was insulin resistant and had high levels of inflammatory markers similar to that of an obese Caucasian.²⁰ The reasons for this susceptibility and the demand for an Asian Indian population-based untargeted metabolomic studies will also be discussed in this review.

Methods

A systematic search of PubMed was conducted to identify the studies that were published up to April 2019. The key term combinations were “metabolomics,” “Type 2 diabetes,” “metabolic profiling,” “prospective studies,” “metabolism,” and “biomarkers.” Additional articles were identified by searching the reference lists from the included studies. Studies were considered probable only if they fulfilled the following criteria: they were prospective-based T2DM studies; metabolomics-based that used the platforms of liquid chromatography—MS , Gas chromatography—MS, or NMR spectroscopy; and examined human samples such as plasma, serum, or urine. The prospective studies, which were based on dietary intake, gestational diabetes, and type 1 diabetes mellitus, were excluded. A total of 172 articles were recognized by our systemic search. A total of 135 articles were omitted based on the predefined inclusion and exclusion criteria. Thus, 35 studies and articles referenced in those studies were included in our review. Even though previously there were prospective-based biomarker discovery studies in T2DM, only in the last decade, studies based on metabolomics were conducted. Hence, articles published from 2010 were included in our review.

In the majority of the prospective studies, the participants were recruited after a questionnaire-based interview. The questionnaire comprised information on sociodemographics, dietary intake, lifestyle, health behavior, medical history, medication, etc. Anthropometric measurements of the participants, such as height, weight, waist–hip ratio, and so on, were determined during the screening process. Most of the studies excluded people with any illness, including diabetes or any other disorders, allergy, those who are undergoing treatment, smokers, and so on. All studies described prediabetes and DM as defined by the American Diabetes Association.²¹

Results

Prospective studies in diabetes research

A majority of patients are asymptomatic until T2DM has reached an advanced stage.²⁵ Thus, it is critical to identify individuals before the onset of T2DM because preventive therapies do exist and also associated complications accumulate over time.¹⁸ To develop preventive strategies, identification of early metabolic alterations is a prerequisite to study etiological pathways and also to recognize high-risk individuals.²⁶ Even though several biomarkers, such as C-reactive protein (CRP), fibrinogen, adiponectin, and so on, are used as indicators, they are not specific to etiology of T2DM.²⁶

Experimental evidences from many prospective studies suggest that blood levels of several metabolites, such as branched-chain amino acids (BCAAs), aromatic amino acids, phospholipids, triglycerides, hexoses, and so on, were associated with the incidence of prediabetes and T2DM.²⁷ Metabolites used in targeted approach were based on prediction using discrimination model such as receiver operating characteristic curve.^24,28,29 The metabolites that are used as biomarkers for T2DM using targeted analysis and its association with T2DM are summarized in Table 1.

Table 1.

List of Metabolites Identified as the Biomarker for Type 2 Diabetes Mellitus Using Targeted Metabolomics-Based Prospective Studies

Metabolites		Prediction	Population	Refs.	No. of participants
BCAAs (isoleucine, leucine, valine), tyrosine, phenylalanine	↑	Prediabetes and type 2 diabetes	American	Wang et al.¹⁸	378 (189—nondiabetic/189—T2DM)
Combination of isoleucine, tyrosine, phenylalanine	↑	Prediabetes and type 2 diabetes	American	Wang et al.¹⁸	378 (189—nondiabetic/189—T2DM)
BCAAs, phenylalanine and α-1-acid glycoprotein, alanine, lactate, pyruvate, and tyrosine	↑	Prediabetes	Finnish	Würtz et al.³⁰	618 (Nondiabetic)
BCAAs (isoleucine, leucine, valine), tyrosine, phenylalanine	↑	Prediabetes and type 2 diabetes	American	Shah et al.³¹	500
			Finnish	Wurtz et al.³²	1680 (Nondiabetic)
			European, South Asian, African Caribbean	Tillin et al.³³	643 South Asian men (416 nondiabetic/227 T2DM) and 801 European men (688 nondiabetic/113 T2DM)
Glutamine	↓	Type 2 diabetes	Finnish	Stancakova et al., 2012⁴⁵	1775 (1624—Nondiabetic/151—T2DM)
Glutamine-glutamate ratio	↓	Type 2 diabetes	American	Cheng et al., 2012⁵¹	601 (Nondiabetic)
Combined glutamine and glutamate	↑	Insulin resistance and impaired glucose tolerance	European American, Hispanic, African American	Palmer et al., 2015³⁵	147 (72—High insulin sensitive/75—low insulin sensitive)
Glycine	↓	Insulin resistance and impaired glucose tolerance	European American, Hispanic, African American	Palmer et al., 2015³⁵	147 (72—High insulin sensitive/75—low insulin sensitive)
2-Aminoadipic acid	↑	Development of type 2 diabetes	American	Wang et al., 2013³⁸	376 (188—Nondiabetic/188—T2DM)
Glycine, LPC	↓	Prediabetes and type 2 diabetes	German	Wang-Sattler et al., 2012³⁶	1297 (1206—Nondiabetic/91—T2DM)
Acetylcarnitine C2	↑	Prediabetes and type 2 diabetes	German	Wang-Sattler et al., 2012³⁶	1297 (1206—Nondiabetic/91—T2DM)
Glycine, lysophosphatidylcholine Acyl (C18:2)	↓	Type 2 diabetes	Korean	Lee et al., 2016³⁷	2240 (924—Normal glucose tolerance/799—prediabetic/517—diabetic)
Phosphatidylcholine acyl-alkyl (C36:0)	↑	Type 2 diabetes	Korean	Lee et al., 2016³⁷
Glycine, serine, LGPC	↓	Impaired glucose tolerance	European	Cobb et al., 2015³⁴	955 (112—Impaired glucose tolerance/843—normal glucose tolerance)
α-HB, α-Ketobutyric acid, 2-aminoadipic acid, oleic acid, 3-MOB, 4-MOP	↑	Impaired glucose tolerance	European	Cobb et al., 2016⁸¹	679 (56—Impaired glucose tolerance/623—normal glucose tolerance)
TAGs of lower carbon number and double bond	↑	Prediabetes and type 2 diabetes	American	Rhee et al.²²	378 (189—Nondiabetic/189—T2DM)
LGPC	↓	Prediabetes	European	Ferrannini et al.²³	1048 (779—Nondiabetic/123—T2DM)
Glycerol, β-HB, and acetoacetate	↑	Prediabetes and type 2 diabetes	Finnish	Mahendran et al., 2013^46,47	4335 (4059—Nondiabetic/276—T2DM)
Hexose, phenylalanine, diacyl-PC (C32:1, C36:1, C38:3, C40:5)	↑	Type 2 diabetes	European	Floegel et al.²⁶	3082 (2282—Nondiabetic/800—T2DM)
Glycine, SMC16:1, LPC C18:2, acyl-alkyl-PC C34:3, C40:6, C42:5, C44:4, C44:5	↓	Type 2 diabetes	European	Floegel et al.²⁶	3082 (2282—Nondiabetic/800—T2DM)
Glyoxylate	↑	Type 2 diabetes	German	Padberg et al.²⁵	177 (121—Prediabetic/56—T2DM)
Alanine, arginine, isoleucine, proline, tyrosine, valine, hexose, PC (C32:1, C34:1, C36:1, C40:5, C42:5)	↑	Type 2 diabetes	Korean	Yang et al., 2019⁵⁴	1939 (1657—Nondiabetic/282—T2DM)
Lyso-phosphatidylcholine acyl C17:0 and C18:2	↓	Type 2 diabetes	Korean	Yang et al., 2019⁵⁴	1939 (1657—Nondiabetic/282—T2DM)
ApoA1, ApoB, LDL, small LDL particles, small VLDL particles, large VLDL particles	↑	Worsening of glycemia and type 2 diabetes	Finnish	Fizelova et al., 2015⁴⁸	6607 (6221—Nondiabetic/386—T2DM)
ApoA1/HDL-C ratio, ApoB/LDL-C ratio, ApoB/non-HDL-C ratio	↑
HDL-C, large HDL particles	↓
Glycerophospholipids and diacylglycerides (C18:2)	↓	Type 2 diabetes	Swedish	Kjellqvist et al., 2016⁸⁵	448 (293—Nondiabetic/155—T2DM)
Palmitoylcarnitine	↑	Type 2 diabetes	Chinese	Qiu et al., 2016³⁹	3118 (1559—Nondiabetic/1559—T2DM)
Free carnitine, medium-chain and long-chain acylcarnitines	↑	Type 2 diabetes	Chinese	Sun et al., 2016⁵³	2103 (1596—Nondiabetic/507—T2DM)
Carnitine precursors, medium-chain dicarboxylic and C12 acylcarnitines	↓	Type 2 diabetes	Chinese	Sun et al., 2016⁵³	2103 (1596—Nondiabetic/507—T2DM)

↑ Represents the positive association, and ↓ represents the negative association.

3-MOB, 3-methyl-2-oxobutyric acid, 4-MOP, 4-methyl-2-oxopentanoic acid; BCAA, branched-chain amino acid; HB, hydroxybutyrate; HDL, high-density lipoprotein; HDL-C, high-density lipoprotein cholesterol; LDL, low-density lipoprotein; LDL-C, low-density lipoprotein cholesterol; LGPC, linoleoylglycerophosphocholine; LPC, lysophosphatidylcholine; T2DM, type 2 diabetes mellitus; TAG, triacylglycerol; VLDL, very low-density lipoprotein.

Several studies included in this review (Table 1) suggest an increase in the levels of BCAAs in prediabetic and diabetic individuals.^{18,30

–34} Most studies showed an inverse association of glycine with prediabetes/T2DM.^26,34

–37 The prediction power of glutamine to glutamate ratio in insulin resistance (IR) was illustrated by Palmer et al.³⁵ An increase in the two metabolites, 2-aminoadipic acid and palmitoylcarnitine, was able to predict the development of T2DM.^38,39 TAGs of lower carbon number and double bond²² and several acylcarnitines were observed to have a significant association with prediabetes and T2DM.^26,35,36

The metabolites identified using untargeted metabolomics-based prospective study that can be used as potential biomarkers and their associations with T2DM are listed in Table 2.

Table 2.

List of Metabolites Identified as the Biomarker for Type 2 Diabetes Mellitus Using Untargeted Metabolomics-Based Prospective Studies

Metabolites		Prediction	Population	Refs.	No. of participants
α-HB	↑	Insulin resistance and impaired glucose regulation	European	Gall et al., 2010⁴²	399 (Nondiabetic)
Hexose sugars, isopentenyladenosine-5-monophosphate	↑	Type 2 diabetes	European	Drogan et al., 2015⁴³	600 (300—Nondiabetic/300—T2DM)
2-Hydroxybiphenyl	↑	Type 2 diabetes	American Indians	Zhao et al., 2015⁴⁴	431 (298—Nondiabetic/133—T2DM)
3-Methyl-2-oxovalerate, BCAA	↑	Impaired glucose tolerance	European	Menni et al., 2013⁴⁰	2204 (1897—Nondiabetic/115—T2DM/192—impaired fasting glucose)
Aminomalonic acid, lysophosphatidylinositol (16:1), urea, d-galactose, palmitic, stearic, oleic, and linoleic acids, BCAAs	↑	Type 2 diabetes	Chinese	Lu et al., 2016⁴¹	394 (197—Nondiabetic/197—T2DM
Proline, glycerol, CMPF, medium chain acylcarnitine (C10)	↓	Type 2 diabetes	Chinese	Lu et al., 2016⁴¹	394 (197—Nondiabetic/197—T2DM
Tyrosine, barogenin	↑	Type 2 diabetes	Swedish	Fall et al., 2016⁸²	4512 (2274—Nondiabetic/1355—impaired fasting glucose/883—diabetic)
LPC (20:2), (18:0)	↓	Type 2 diabetes	Swedish	Fall et al., 2016⁸²
α-HB, glucose, mannose	↑	Type 2 diabetes	Finnish	Peddinti et al., 2017⁸³	543 (397—Nondiabetic/146—diabetic)
α-Tocopherol, bradykinin hydroxyproline	↓	Type 2 diabetes	Finnish	Peddinti et al., 2017⁸³	543 (397—Nondiabetic/146—diabetic)
LPC (19:1), phosphatidylcholine (17:0/18:2), 2-Hydroxyethanesulfonate	↓	Type 2 diabetes	Swedish	Shi et al., 2018⁸⁴	1006 (503—Nondiabetic/503—T2DM)
Deoxycholic, glycodeoxycholic acid, BCAAs	↑	Type 2 diabetes	African	Zeng et al., 2019⁴⁹	75 (28—Nondiabetic/20—T2DM/27—impaired glucose tolerance)
LPC (18:2)	↓	Type 2 diabetes	African	Zeng et al., 2019⁴⁹	75 (28—Nondiabetic/20—T2DM/27—impaired glucose tolerance)
3-(4-Hydroxyphenyl)lactate, erythritol, BCAAs	↑	Type 2 diabetes	American	Rebholz et al., 2018⁵²	1006 (503—Nondiabetic/503—T2DM)
Asparagine	↓	Type 2 diabetes	American	Rebholz et al., 2018⁵²	1006 (503—Nondiabetic/503—T2DM)
Tyrosine, alanine, isoleucine, aspartate, glutamate	↑	Type 2 diabetes	Finnish	Vangipurapu et al., 2019⁵⁰	4851 (4329—Nondiabetic/522—T2DM)

↑ Represents the positive association, and ↓ represents the negative association.

CMPF, 3-carboxy-4-methyl-5-propyl-2-furanpropionic acid.

Similar to the targeted metabolomics-based studies, some of the studies using untargeted metabolomics reported an increase in the levels of BCAAs in individuals with prediabetes and T2DM.^40,41 An increase in α-hydroxybutyrate (α-HB) levels was able to predict insulin resistance and impaired glucose regulation.⁴² In addition, hexose sugars were found to have a strong association with T2DM.⁴³

The number of prospective studies using untargeted metabolomics was comparatively less than targeted metabolomics-based prospective studies. Most of the studies had the elderly population as their study cohort, while only two studies had a younger population.^32,44 Nine of the studies had a single-gender cohort,^{33,40,41,45

–50} whereas the rest included both men and women. The majority of the cohorts were of European descent, eight of American,^{18,22,31,35,36,38,51,52} three of Chinese,^39,41,53 two of Korean,^37,54 one each of American Indians,⁴⁴ and African.⁴⁹ Only three studies had multiethnic participants.^31,33,35 Out of the 11 untargeted metabolomics-based prospective studies, 9 had both men and women and in that only 1 had a younger group of study participants.⁴⁴ The study by Zhao et al.,⁴⁴ had a greater advantage since it was untargeted, gender-balanced, and had young participants. The only drawback was that it had only a single ethnic group so generalization to other groups could only be done with caution.

Discussion

A series of events leads to the development of T2DM. The pathogenesis of T2DM is characterized by a triad of genetic predisposition, environment, and acquired organ dysfunction. The genetic predisposition influenced by environmental factors such as high-calorie foods, low energy expenditure, and so on, along with beta-cell dysfunction, insulin resistance, and hepatic glucose production disrupt the glucose homeostasis system leading to prediabetes and T2DM.⁵⁵ Insulin resistance and beta-cell dysfunction are the initial decisive pathogenic events that cause the transition of normal glucose tolerance to impaired glucose tolerance leading to T2DM. Hence, routine biomarkers of T2DM such as glucose reflect only at the later stages of disease pathogenesis. However, metabolites identified from metabolomics-based prospective studies can reveal the early stages of the disease progression.

The role of metabolites in regulating insulin signaling is elucidated by huge data sets produced by metabolomics analysis and this had contributed to the evidence that insulin resistance is triggered by the intricate interaction of numerous metabolic pathways. In the review by Yang et al., how metabolites such as lipids, amino acids, and bile acids modulate different components in the insulin signaling pathway is meticulously explained.⁵⁶ Increased lipid deposition in adipose tissue leads to activation of pro-inflammatory cytokines, which in turn increases adipose lipolysis. The suppression of lipolysis by insulin is disrupted in the case of insulin resistance. Therefore, lipids play a key role in metabolic inflexibility and insulin resistance.

Several lipids such as saturated fatty acids (SFAs), diacylglycerols (DAGs), and ceramides influence insulin resistance, while some lipids such as poly- (PUFA) and mono-unsaturated fatty acids (MUFA) improve insulin sensitivity. SFAs activate pro-inflammatory signaling pathways, alter membrane distribution of tyrosine protein kinase sarcoma (SRC), and induce endoplasmic reticulum stress, thus intensifying insulin resistance. Even though PUFA and MUFA show a positive association with insulin sensitivity in animal studies, results are not conclusive in human studies.⁴¹

Although triacylglycerols (TAGs) are found to be elevated in T2DM, there is no evidence suggesting its role in insulin resistance. DAGs, the immediate precursor of TAGs, activate protein kinase C (PKC) isoforms that alternately cause insulin resistance by modifying phosphorylation of insulin signaling molecules. Ceramides, lipids composed of sphingosine and fatty acids, are associated with insulin resistance. Ceramides activate PKC, which prevents AKT binding to phosphatidylinositol (3,4,5)-trisphosphate (PIP₃), thus interfering in insulin signaling. In animal-based studies, treating obese mice with specific phospholipids such as PC (C18:0/C18:1) showed improvement in insulin sensitivity and glucose tolerance.⁵⁶

Elevation of BCAAs leads to accumulation of acylcarnitines in muscles, which induce oxidative stress and mitochondrial dysfunction, thereby worsening insulin sensitivity, while glycine, on binding to acyl-CoA, reduces the accumulation of acylcarnitines in the muscles.⁵⁶ The elevation of BCAAs before the onset of the disease in most of the studies may substantiate it to be a potential biomarker of T2DM.

An ideal example of a prospective metabolomic study is the metabolic profiling of pregnant women to identify women at risk for gestational diabetes mellitus (GDM). The study is regarded as prospective as the women will be followed throughout the pregnancy. The participants were recruited during their first trimester, and samples were also collected. The participants were categorized as GDM or normal glucose tolerant in their third trimester on the basis of their oral glucose tolerance test (OGTT) results.⁵⁷ An advantage is that the samples can be collected at various time points when the women report for their routine medical examination. Early detection of GDM is necessary to enhance the well-being of both mother and fetus, as well as to reduce the risk of developing T2DM.⁵⁷

As GDM shares a resemblance with the pathophysiology of T2DM, biomarkers predictive for T2DM can be used for targeted metabolomic analyses of GDM. However, the prognostic power of these biomarkers may not be the same in GDM.⁵⁷ Few metabolites exhibited discrepancy across studies, owing to diverse study methods and varying sample sizes.⁵⁸ Metabolites that showed such inconsistency were BCAAs; some studies demonstrated elevated levels of BCAAs, while others exhibited no such correlation.^58

–62 A metabolomics-based GDM longitudinal study revealed that there was a distinct separation between GDM women who were euglycemic and who developed T2DM postpartum.⁶³ In addition, elevated levels of 2-HB and 3-HB were associated with an increased risk of developing T2DM. These could be deliberated as potential biomarkers for early onset of T2DM in GDM women. Therefore, similar studies are necessary for additional insight into identifying a potential biomarker for GDM.

According to the International Diabetes Federation (IDF) in 2017, out of 425 million people having diabetes in the world, 82 million people are in the South East Asia region.⁶⁴ According to IDF estimates in 2017, the second largest number of adults living with diabetes worldwide was India, after China.

Urbanization and its associated lifestyle changes had led to a rapid epidemic of T2DM.⁶⁵ Furthermore, Indians make up specific phenotype, which renders them more susceptible to T2DM. This phenotype was referred to as Asian Indian Phenotype.⁶⁶ The Asian Indian phenotype demonstrated lower thresholds for risk factors such as age, obesity, abdominal adiposity, and body fat content.⁶⁷ Indians developed T2DM at a very young age and showed a higher age-related prevalence of T2DM compared to Caucasians.^68,69 Furthermore, healthy nonobese Asian Indians had higher plasma concentrations of nonesterified fatty acids, leptin, CRP, and fasting plasma insulin and a lower concentration of adiponectin similar to that of an obese Caucasian.^20,70 Indians had a lower BMI with higher central adiposity.^71,72 The risk for T2DM was more if BMI exceeds 23 kg/m². The higher body fat and lower muscle mass could be the rationale for increased insulin resistance in Indians.⁷¹

One reason for the susceptibility to T2DM could be due to the consumption of low-energy vegetarian diet by Indians for centuries instead of the high-energy meat diet in Europe.⁷³ Another reason could be a gene adaptation strategy, which was requisite to efficiently store all energy at the time of high abundance of food to survive a prolonged period of starvation. This genotype was known as “Thrifty.”^65,73

–76 Both of these explanations might be the underlying causes of this kind of vulnerability. Also with urbanization, consumption of high energy food increased and the energy expenditure minimized, which led to an overall surge in obesity and T2DM.⁶⁵

In India, there are many metabolomics-based studies for biomarkers of T2DM and prediabetes; for example, vitamin D metabolites, SFAs, lactate, valine, isoleucine, and phenylalanine were found to be potential biomarkers for T2DM.^77,78 But there are no metabolomics-based prospective studies in India with the susceptible Indian population as their study cohort. Of the metabolomics-based prospective studies involving Indians, a majority of them had Indians living abroad as their study cohort. For instance, a prospective study conducted on American Indians could identify elevated levels of 2-hydroxybiphenyl and reduced levels of glycerophospholipids significantly associated with an increased risk of diabetes.⁴⁴ However, those metabolites identified could not be generalized to the Asian Indian population due to a difference in their dietary habits and lifestyle. The thrifty genotype, dietary habits, related microbiota, and lack of appropriate studies demand untargeted metabolomic profiling based prospective study in India to identify a potential biomarker and also to understand the population-specific disease mechanism.

Conclusion

Untargeted metabolomic profiling facilitates exploration of a large portion of the metabolome, which in turn leads to the discovery of novel metabolites and utilization of these metabolites as biomarkers for disease. A biomarker should be able to point out the disruption in the pathway before the disease onset but it is not necessary that the metabolite should be in the pivotal pathway.⁷⁹ As mentioned, metabolic diseases are often latent for a long time before becoming clinically evident; therefore, a prospective study will help in monitoring these metabolic alterations beforehand.

A good number of the previous prospective studies had an elderly population as their cohort; for that reason, caution should be taken before generalization to other population. Some studies had single gender as their cohort; consequently, one should understand that the metabolites could be biased by gender-specific hormones.⁴⁰ The majority of studies were done in Caucasians; therefore, the whole disease risk in Asian population could not be identified. Aspects such as geographic diverseness, genetic susceptibility, lifestyle factors, and dietary habits cause a large degree of dissimilarity between the Asian population and the Caucasians. As the second largest number of adults living with diabetes worldwide is in India, there is a dire need for a prospective study of young Asian Indians with an equal proportion of men and women.

Metabolites such as BCAA that are strong predictors of T2DM were recently identified to have a positive association with incident cardiovascular disease (CVD).⁸⁰ Hence, the discovery of new biomarkers for T2DM could also be used as biomarkers for CVD. Genetic analysis along with the metabolomics could be used for the discovery of a new pathophysiological mechanism of T2DM. Traditional risk factors and Anthropometry/Biochemical measures along with the discovered predictive metabolites could make a powerful discrimination model for T2DM.

Ethical Approval

This review was conducted in accordance with ethical standard.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the Department of Biotechnology, Ministry of Science and Technology, government of India (BT/PR8444/MED/30/1021/2013).

References

Hollywood

, Brison

, Goodacre

. Metabolomics: Current technologies and future trends. Proteomics, 2006; 6:4716–4723.

Kosmides

, Kamisoglu

, Calvano

, et al. Metabolomic fingerprinting: Challenges and opportunities. Crit Rev Biomed Eng, 2013; 41:205–221.

Schmidt

. Metabolomics: What's happening downstream of DNA. Environ Health Perspect, 2004; 112:A410–A415.

Goodacre

, Vaidyanathan

, Dunn

, et al. Metabolomics by numbers: Acquiring and understanding global metabolite data. Trends Biotechnol, 2004; 22:245–252.

Barabasi

A-L

, Oltvai

ZNZN

, Barabási

A-L

. Network biology: Understanding the cell's functional organization. Nat Rev Genet, 2004; 5:101–113.

Preet

. Metabolomics: Approaches and applications to diabetes research. J Diabetes Metab, 2013; 01(S6). DOI:10.4172/2155-6156.S6-001

Powers AC. Diabetes mellitus: Diagnosis, classification, and pathophysiology. In: Kasper

, Fauci

, Hauser

, Longo

, Jameson

, Loscalzo

(eds). Harrison's Principles of Internal Medicine. McGraw-Hill Medical; 2014. http://aplicacionesbiblioteca.udea.edu.co:2351/content.aspx?sectionid=79752868&bookid=1130&Resultclick=2&q=diabetes+mellitus%3A+diagnosis%2C+classification%2C+and+pathophysiology

Friedrich

. Metabolomics in diabetes research. J Endocrinol, 2012; 215:29–42.

Lakhtakia

. The history of diabetes mellitus. Sultan Qaboos Univ Med J, 2013; 13:368–370.

10.

, Xie

, Jia

, et al. Metabolomics in human type 2 diabetes research. Front Med, 2013; 7:4–13.

11.

Salek

, Maguire

. A metabolomic comparison of urinary changes in type 2 diabetes in mouse, rat, and human. Physiol Genomics, 2007; 29:99–108.

12.

Fiehn

, Timothy Garvey

, Newman

, et al. Plasma metabolomic profiles reflective of glucose homeostasis in non-diabetic and type 2 diabetic obese African-American women. PLoS One, 2010; 5:1–10.

13.

Huffman

, Shah

, Stevens

, et al. Relationships between circulating metabolic intermediates and insulin action in overweight to obese, inactive men and women. Diabetes Care, 2009; 32:1678–1683.

14.

Newgard

, An

, Bain

, et al. A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance. Cell Metab, 2009; 9:311–326.

15.

Oberbach

, Blüher

, Wirth

, et al. Combined proteomic and metabolomic profiling of serum reveals association of the complement system with obesity and identifies novel markers of body fat mass changes. J Proteome Res, 2011; 10:4769–4788.

16.

Adams

, Pratipanawatr

, Berria

, et al. Ceramide content is increased in skeletal muscle from obese insulin-resistant humans. Diabetes, 2004; 53:25–31.

17.

Mihalik

, Michaliszyn

, De Las Heras

, et al. Metabolomic profiling of fatty acid and amino acid metabolism in youth with obesity and type 2 diabetes: Evidence for enhanced mitochondrial oxidation. Diabetes Care, 2012; 35:605–611.

18.

Wang

, Larson

, Vasan

, et al. Metabolite profiles and the risk of developing diabetes. Nat Med, 2011; 17:448–453.

19.

Tabák

, Jokela

, Akbaraly

, et al. Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: An analysis from the Whitehall II study. Lancet (London, England), 2009; 373:2215–2221.

20.

Abate

, Chandalia

, Snell

, et al. Adipose tissue metabolites and insulin resistance in nondiabetic Asian Indian men. J Clin Endocrinol Metab, 2004; 89:2750–27505.

21.

American Diabetes Association. 2. Classification and diagnosis of diabetes. Diabetes Care, 2017; 40(Suppl 1):S11–S24.

22.

Rhee

, Cheng

, Larson

, et al. Lipid profiling identifies a triacylglycerol signature of insulin resistance and improves diabetes prediction in humans. J Clin Invest, 2011; 121:1402–1411.

23.

Ferrannini

, Natali

, Camastra

, et al. Early metabolic markers of the development of dysglycemia and type 2 diabetes and their physiological significance. Diabetes, 2013; 62:1730–1737.

24.

Wilson

PWF

, Meigs

, Sullivan

, et al. Prediction of incident diabetes mellitus in middle-aged adults: The Framingham Offspring Study. Arch Intern Med, 2007; 167:1068–1074.

25.

Padberg

, Peter

, González-Maldonado

, et al. A new metabolomic signature in type-2 diabetes mellitus and its pathophysiology. PLoS One, 2014; 9:e85082.

26.

Floegel

, Stefan

, Yu

, et al. Identification of serum metabolites associated with risk of type 2 diabetes using a targeted metabolomic approach. Diabetes, 2013; 62:639–648.

27.

Guasch-Ferré

, Hruby

, Toledo

, et al. Metabolomics in prediabetes and diabetes: A systematic review and meta-analysis. Diabetes Care, 2016; 39:833–846.

28.

Schulze

, Weikert

, Pischon

, et al. Use of multiple metabolic and genetic markers to improve the prediction of type 2 diabetes: The EPIC-Potsdam study. Diabetes Care, 2009; 32:2116–2119.

29.

Schulze

, Hoffmann

, Boeing

, et al. An accurate risk score based on anthropometric, dietary, and lifestyle factors to predict the development of type 2 diabetes. Diabetes Care, 2007; 30:510–515.

30.

Würtz

, Tiainen

, Makinen

, et al. Circulating metabolite predictors of glycemia in middle-aged men and women. Diabetes Care, 2012; 35:1749–1756.

31.

Shah

, Crosslin

, Haynes

, et al. Branched-chain amino acid levels are associated with improvement in insulin resistance with weight loss. Diabetologia, 2012; 55:321–330.

32.

Wurtz

, Soininen

, Kangas

, et al. Branched-chain and aromatic amino acidsare predictors of insulinresistance in young adults. Diabetes Care, 2013; 36:648–655.

33.

Tillin

, Hughes

, Wang

, et al. Diabetes risk and amino acid profiles: Cross-sectional and prospective analyses of ethnicity, amino acids and diabetes in a South Asian and European cohort from the SABRE (Southall And Brent REvisited) Study. Diabetologia, 2015; 58:968–979.

34.

Cobb

, Eckhart

, Perichon

, et al. A novel test for IGT utilizing metabolite markers of glucose tolerance. J Diabetes Sci Technol, 2015; 9:69–76.

35.

Palmer

, Stevens

, Antinozzi

, et al. Metabolomic profile associated with insulin resistance and conversion to diabetes in the Insulin Resistance Atherosclerosis Study. J Clin Endocrinol Metab, 2015; 100:E463–E468.

36.

Wang-Sattler

, Yu

, Herder

, et al. Novel biomarkers for pre-diabetes identified by metabolomics. Mol Syst Biol, 2012; 8:615.

37.

Lee

, Xu

, Lee

, et al. Identification of putative biomarkers for type 2 diabetes using metabolomics in the Korea Association REsource (KARE) cohort. Metabolomics, 2016. DOI:10.1007/s11306-016-1103-9

38.

Wang

, Ngo

, Psychogios

, et al. 2-Aminoadipic acid is a biomarker for diabetes risk. J Clin Invest, 2013; 123:4309–4317.

39.

Qiu

, Zheng

, Wang

, et al. Plasma metabolomics identified novel metabolites associated with risk of type 2 diabetes in two prospective cohorts of Chinese adults. Int J Epidemiol, 2016; 45:1507–1516.

40.

Menni

, Fauman

, Erte

, et al. Biomarkers for type 2 diabetes and impaired fasting glucose using a nontargeted metabolomics approach. Diabetes, 2013. 62:4270–4276.

41.

, Wang

, Ong

C-N

, et al. Metabolic signatures and risk of type 2 diabetes in a Chinese population: An untargeted metabolomics study using both LC-MS and GC-MS. Diabetologia, 2016; 59:2349–2359.

42.

Gall

, Beebe

, Lawton

, et al. α-hydroxybutyrate is an early biomarker of insulin resistance and glucose intolerance in a nondiabetic population. PLoS One, 2010; 5:e10883.

43.

Drogan

, Dunn

, Lin

, et al. Untargeted metabolic profiling identifies altered serum metabolites of type 2 diabetes mellitus in a prospective, nested case control study. Clin Chem, 2015; 61:487–497.

44.

Zhao

, Zhu

, Hyun

, et al. Novel metabolic markers for the risk of diabetes development in American Indians. Diabetes Care, 2015; 38:220–227.

45.

Stančáková

, Civelek

, Saleem

, et al. Hyperglycemia and a common variant of GCKR are associated with the levels of eight amino acids in 9,369 Finnish men. Diabetes, 2012; 61:1895–1902.

46.

Mahendran

, Cederberg

, Vangipurapu

, et al. Glycerol and fatty acids in serum predict the development of hyperglycemia and type 2 diabetes in Finnish men. Diabetes Care, 2013; 36:3732–3738.

47.

Mahendran

, Vangipurapu

, Cederberg

, et al. Association of ketone body levels with hyperglycemia and type 2 diabetes in 9,398 Finnish men. Diabetes, 2013; 62:3618–3626.

48.

Fizelova

, Miilunpohja

, Kangas

, et al. Associations of multiple lipoprotein and apolipoprotein measures with worsening of glycemia and incident type 2 diabetes in 6607 non-diabetic Finnish men. Atherosclerosis, 2015; 240:272–277.

49.

Zeng

, Mtintsilana

, Goedecke

, et al. Alterations in the metabolism of phospholipids, bile acids and branched-chain amino acids predicts development of type 2 diabetes in black South African women: A prospective cohort study. Metabolism, 2019; 95:57–64.

50.

Vangipurapu

, Stancáková

, Smith

, et al. Nine amino acids are associated with decreased insulin secretion and elevated glucose levels in a 7.4-year follow-up study of 5,181 Finnish men. Diabetes, 2019. DOI:10.2337/db18-1076

51.

Cheng

, Rhee

, Larson

, et al. Metabolite profiling identifies pathways associated with metabolic risk in humans. Circulation, 2012; 125:2222–2231.

52.

Rebholz

, Yu

, Zheng

, et al. Serum metabolomic profile of incident diabetes. Diabetologia, 2018; 61:1046–1054.

53.

Sun

, Liang

, Gao

, et al. Early prediction of developing type 2 diabetes by plasma acylcarnitines: A population-based study. Diabetes Care, 2016; 39:1563–1570.

54.

Yang

, Kwak

, Jo

, et al. Serum metabolite profile associated with incident type 2 diabetes in Koreans: Findings from the Korean Genome and Epidemiology Study. Sci Rep, 2018; 8:8207.

55.

Leahy

. Pathogenesis of type 2 diabetes mellitus. Arch Med Res, 36:197–209.

56.

Yang

, Vijayakumar

, Kahn

. Metabolites as regulators of insulin sensitivity and metabolism. Nat Rev Mol Cell Biol, 2018; 19:654–672.

57.

Bentley-Lewis

, Huynh

, Xiong

, et al. Metabolomic profiling in the prediction of gestational diabetes mellitus. Diabetologia, 2015; 58:1329–1332.

58.

Huynh

, Xiong

, Bentley-Lewis

. A systematic review of metabolite profiling in gestational diabetes mellitus. Diabetologia, 2014; 57:2453–2464.

59.

Scholtens

, Bain

, Reisetter

, et al. Metabolic networks and metabolites underlie associations between maternal glucose during pregnancy and newborn size at birth. Diabetes, 2016; 65:2039–2050.

60.

Newgard

. Metabolomics and metabolic diseases: Where do we stand?. Cell Metab, 2017; 25:43–56.

61.

Allalou

, Nalla

, Prentice

, et al. A predictive metabolic signature for the transition from gestational diabetes mellitus to type 2 diabetes. Diabetes, 2016; 65:2529–2539.

62.

Scholtens

, Muehlbauer

, Daya

, et al. Metabolomics reveals broad-scale metabolic perturbations in hyperglycemic mothers during pregnancy. Diabetes Care, 2014; 37:158–166.

63.

Dudzik

, Zorawski

, Skotnicki

, et al. GC–MS based gestational diabetes mellitus longitudinal study: Identification of 2-and 3-hydroxybutyrate as potential prognostic biomarkers. J Pharm Biomed Anal, 2017; 144:90–98.

64.

International Diabetes Federation. IDF Diabetes Atlas—Home. 8th ed. 2017. www.diabetesatlas.org

65.

Shah

, Mohan

. Diabetes in India: What is different?. Curr Opin Endocrinol Diabetes Obes, 2015; 22:283–289.

66.

Mohan

, Sandeep

, Deepa

, et al. Epidemiology of type 2 diabetes: Indian scenario. Indian J Med Res, 2007; 125:217–230.

67.

Ramachandran

, Snehalatha

, Vijay

. Low risk threshold for acquired diabetogenic factors in Asian Indians. Diabetes Res Clin Pract, 2004; 65:189–195.

68.

Harris

, Flegal

, Cowie

, et al. Prevalence of diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults. The Third National Health and Nutrition Examination Survey, 1988–1994. Diabetes Care, 1998; 21:518–524.

69.

Group TD-DS. Age, body mass index and type 2 diabetes? Associations modified by ethnicity. Diabetologia, 2003; 46:1063–1070.

70.

McKeigue

, Shah

, Marmot

. Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet, 1991; 337:382–386.

71.

Banerji

, Faridi

, Atluri

, et al. Body composition, visceral fat, leptin, and insulin resistance in Asian Indian men. J Clin Endocrinol Metab, 1999; 84:137–144.

72.

Shelgikar

, Yajnik

, Hockaday

TDR

. Central rather than generalized obesity is related to hyperglycaemia in Asian Indian Subjects. Diabet Med, 1991; 8:712–717.

73.

Nair

, Bigelow

, Asmann

, et al. Asian Indians have enhanced skeletal muscle mitochondrial capacity to produce ATP in association with severe insulin resistance. Diabetes, 2008; 57:1166–1175.

74.

Neel

. Diabetes Mellitus: A “Thrifty” genotype rendered detrimental by “Progress”?. Am J Hum Genet, 1962; 14:353–362.

75.

Hales

, Barker

DJP

. Type 2 (non-insulin-dependent) diabetes mellitus: The thrifty phenotype hypothesis. Int J Epidemiol, 2013; 42:1215–1222.

76.

Yajnik

. Early life origins of insulin resistance and type 2 diabetes in India and other Asian countries. J Nutr, 2004; 134:205–210.

77.

Gogna

, Krishna

, Oommen

, et al. Investigating correlations in the altered metabolic profiles of obese and diabetic subjects in a South Indian Asian population using an NMR-based metabolomic approach. Mol Biosyst, 2015; 11:595–606.

78.

Adela

, Borkar

, Bhandi

, et al. Lower vitamin D metabolites levels were associated with increased coronary artery diseases in type 2 diabetes patients in India. Sci Rep, 2016; 6:37593.

79.

Liu

, Semiz

, van der Lee

, et al. Metabolomics based markers predict type 2 diabetes in a 14-year follow-up study. Metabolomics, 2017; 13:104.

80.

Tobias

, Lawler

, Harada

, et al. Circulating branched-chain amino acids and incident cardiovascular disease in a prospective cohort of US Women. Circ Genomic Precis Med, 2018; 11:e002157.

81.

Cobb

, Eckhart

, Motsinger-Reif

, Carr

, Groop

, Ferrannini

. α-Hydroxybutyric acid is a selective metabolite biomarker of impaired glucose tolerance. Diabetes Care, 2016. DOI:10.2337/dc15-2752.

82.

Fall

, Salihovic

, Brandmaier

, et al. Non-targeted metabolomics combined with genetic analyses identifies bile acid synthesis and phospholipid metabolism as being associated with incident type 2 diabetes. Diabetologia, 2016. DOI:10.1007/s00125-016-4041-1.

83.

Peddinti

, Cobb

, Yengo

, et al. Early metabolic markers identify potential targets for the prevention of type 2 diabetes. Diabetologia, 2017. DOI:10.1007/s00125-017-4325-0.

84.

Shi

, Brunius

, Lehtonen

, et al. Plasma metabolites associated with type 2 diabetes in a Swedish population: a case-control study nested in a prospective cohort. Diabetologia, 2018. DOI:10.1007/s00125-017-4521-y.

85.

Kjellqvist

, Klose

, Surma

, et al. Identification of shared and unique serum lipid profiles in diabetes mellitus and myocardial infarction. J Am Heart Assoc, 2016. DOI:10.1161/JAHA.116.004503.