Table 1.

Metabolomic studies about obesity, insulin resistance or type 2 diabetes mellitus (T2DM) in children.

Methodology	Instrumental Analysis	Disease	Study Design	Sample	Findings	Ref.
Untargeted	LC-MS, CE-MS, GC-MS	Obesity and IR	Fingerprinting study: 60 prepubertal obese children. Boys (n = 30, 50% IR and 50% non-IR) Girls (n = 30, 50% IR and 50% non-IR) Validation study: 100 prepubertal obese children. Boys (n = 50, 50% IR and 50% non-IR) Girls (n = 50, 50% IR and 50% non-IR)	Serum	IR vs non-IR:	[51]
					Inflammation, central carbon metabolism and gut microbiota are the most altered processes. Increased BCAA, ArAAs, Ala, Pro, Pyr, taurodeoxycholate, glycodeoxycholate, piperidine, pyroglutamate. In females, increased free carnitine, propionylcarnitine and butyrylcarnitine, but in males only propionylcarnitine.
Untargeted	LC-MS/MS	Metabolic Risk	Boys (n = 113) Girls (n = 125) (8–14 years)	Serum	Metabolic Risk: In girls:	[54]
					Positive association of DG(16:0/16:0), 1,3-dielaidin, myo-inositol, and urate. Inverse association of thymine, dodecenedioic acid, and n-acetylglycine with metabolic risks.
					In boys::
					Positive associations of BCAA, DG(16:0/16:0), tyrosine, and 5′-methylthioadenosine.
Untargeted	NMR	IR	Cross sectional study: 78 non diabetic adolescents (8–18 years) Longitudinal study: 16 subjects after a mean follow-up of 2.3 years	Plasma	Higher baseline 2-hydroxybutyrate and BCAA levels in insulin resistant youth and predict worsening of glycemic control Alterations of 2-hydroxybutyrate metabolism predict incipient deterioration of β-cell function and longitudinal worsening of glycemic tolerance.	[63]
Untargeted	NMR	IR	170 healthy normal weight children (5, 14 and 16 years)	Serum	IR higher in girls than in boys. In healthy normal weight children IR was associated with reduced concentrations of BCAA, 2-ketobutyrate, citrate and 3-hydroxybutyrate, and higher concentrations of lactate and alanine.	[64]
Semi-targeted	LC-MS/MS, GC-MS	Obesity	Obese (n = 84) Overweight (n = 28) Normal-weight (n = 150) Median age 7.7 years 50% boys 50% girls	Plasma	OB vs NW:	[53]
					Increased BCAA (Val, Leu, Ile) and androgen hormones (DHEA-S).
Semi-targeted	LC-MS/MS, GC-MS	Obesity and IR	Hispanic children Obese (n = 450) Non-obese (n = 353) Boys (n= 405) Girls (n = 398). (4–19 years). Mean age 11.1 years	Plasma	OB vs NOB:	[56]
					Increased BCAA and acylcarnitine catabolism and changes in nucleotides, lysolipids, steroid derivatives and inflammation markers. Reduced fatty acid catabolism. BCAAs, ArAAs, aspartate, dipeptides, citrate, asparagine, glycine and serine is associated with risk factors for IR, hyperleptinemia, hypertriglyceridemia, hyperuricemia and inflammation.
Semi-targeted	LC-MS/MS, GC-MS	Obesity	Longitudinal study for 5 years: Obese (n = 68) Overweight (n = 23) Normal weight (n = 122) 48.8% boys Median age 7.7 years	Plasma	BCAA is not associated with worsening metabolic health during early adolescence. Inverse association of the BCAA pattern with a change in fasting glucose in boys. Direct relation of BCAA pattern with a change in serum triglycerides in girls. Higher score for androgen hormone pattern at baseline corresponds with a decrease in leptin an increase in CRP in girls.	[55]
Targeted	LC-MS/MS, GC-MS	Obesity and IR	100 prepubertal obese children. Boys (n = 50, 50% IR and 50% non-IR) Girls (n = 50, 50% IR and 50% non-IR) 5–10 years	Serum	IR vs non-IR:	[62]
					Higher ALT, GPT and TAG levels Higher leptin and reduce leptin/adiponectin ratio Increase BCAA, ArAAs (Phe, Tyr and Trp), and Ala The most altered pathway is the urea cycle, alanine metabolism and the glucose-alanine cycle. C12 acylcarnitine and methionine correlate with HOMA-IR exclusively in males
Targeted	MS/MS	Obesity and T2D	Case-control: Obese (n = 64) Obese with TD2 (n = 17) Normal-weight (n = 39) 12–17 years	Plasma	T2D vs OB/NW:	[87]
					Decreased BCAA.
					T2D/OB vs NW:
					No difference in long-chain AcylCN Reduced short and medium-chain AcycCN No defects in fatty acid or amino acid metabolism
					No differences in fasting FFA levels
Targeted	MS/MS	Obesity, IR and T2D	Case-control: Obese (n = 57) Obese prediabetes (n = 27) Obese T2D (n = 17) Normal-weight (n = 38) 13–14 years	Plasma	BCAA and BCAA intermediates correlated: positively with insulin sensitivity and DI	[79]
Targeted	LC-MS/MS	Obesity and IR	Cross sectional study: 69 healthy children and adolescents 8-18 years Longitudinal cohort study in subset: Subgroup of 17 participants 8-13 years	Plasma	OB vs NW:	[72]
					Increased BCAA Increased BCAA not associated with measures of insulin resistance at baseline. Baseline BCAAs predicted HOMA-IR at 18 months. Elevations in the concentrations of BCAAs were associated with reduced insulin sensitivity at 12 months.
Targeted	MS/MS	Obesity and IR	Cross-sectional study: Obese (n = 82) Boys (n = 41) Girls (n = 41) 12–18 years	Plasma	BCAA levels and by products of BCAA catabolism are higher in males than females with similar BMI. In males, HOMA-IR correlated:	[74]
					Positively: BMI z-score, BCAA, uric acid, long-chain acyl-carnitines Negatively: fatty-acid oxidation products
					In females, HOMA-IR correlated:
					Positively: BMI z-score
					Adiponectin correlated inversely with BCAA and uric acid in males, but not females
Targeted	LC-MS/MS	Obesity and IR	Identify biomarkers predictive of future disease risk- Obese (n = 46) Obese to normal weight (n = 18) Normal-weight (n = 45) 9–11 years	Plasma	Baseline BCAA concentration as a predictor of future risk of insulin resistance and metabolic syndrome OB vs NW:	[85]
					Increased levels of BCAA, Tyr, Phe, 2-AAA and several acyl-carnitines Lower levels of acyl-alkyl phosphatidylcholines
Targeted	MS/MS	Obesity and IR	Longitudinal study: 80 obese Caucasian children. 40 participate in one-year lifestyle interventions 8–15 years	Serum	Tyr was the only metabolite significantly associated with HOMA-IR at baseline and after 1-year intervention. No association between HOMA-IR and BCAA.	[84]
Targeted	MS/MS	Obesity and IR	430 control (13–15 years). 91 morbid obese (12–16 years)	Plasma	Accumulation of ADMA is associated with modulation of insulin signaling and insulin resistance. ADMA decreased after obesity intervention program	[68]
Targeted	MS/MS, LC-MS/MS	Obesity and IR	Meta-analysis 1020 pre-pubertal children from three European studies. 8–10 years	Plasma	Positive association of SM (32:2) with BMI z-score. SM 32:2 as a potential molecular marker for mechanistic alterations involved in the pathogenesis of obesity. Ala and Tyr was associated positively with HOMA-IR. Acylcarnitines and non-esterified fatty acids were negatively associated with HOMA.	[67]
Targeted	GC-MS	Obesity and IR	20 obese with IR 67 obese without IR 8.5–17.9 years	Urine	The steroidal signature IR vs non-IR:	[59]
					High adrenal androgens, glucocorticoids and mineralocorticoid metabolites Higher 5α-reductase and 21-hidroxylase activity Lower 11bHSD1 activity
					The authors suggest a vicious cycle model, whereby glucocorticoids induce IR.
Targeted	MS/MS	Obesity and Metabolic Risk	Non-OW/OB and low MetRisk (n = 335) Non-OW/OB and high MetRisk (n = 29) OW/OB and low MetRisk (n = 58) OW/OB and high MetRisk (n = 102) Girls 48.3% Boys 51.7% 11–16 years	Plasma	Lower levels of LCFA in non-OW/OB with high MetRisk and OW/OB with high MetRisk compared to mon-OW/OB with low MetRisk. Higher levels of BCAA metabolite pattern in OW/OB with high MetRisk compared to non-OW/OB with low MetRisk. Higher levels of DAG in OW/OB with high MetRisk vs non-OB/OW with low MetRisk. Higher score of androgen steroid hormones pattern in OW/OB with high MetRisk compared to Non-OW/OB with low MetRisk. Higher levels of AcylCN in non-OW/OB with high MetRisk compared to non-OW/OB with low MetRisk. Lower levels of AcylCN in OW/OB with high MetRisk compared to Non-OW/OB with low MetRisk.	[76]

Abbreviations: 11bHSD1: 11β-hydroxysteroid dehydrogenase type 1; 2-AAA: alpha amino adipic acid; AcylCN: acylcarnitines; ADMA: asymmetric dimethylarginine; Ala: alanine; ALT: alanine transaminase; ArAAs: aromatic amino acids; BCAA: branched chain amino acids; BMI: body mass index; CE-MS: capillary electrophoresis – mass spectrometry; CRP: C-reactive protein; DAG: diacylglycerides; DG: diglyceride; DI: disposition index; GC-MS: gas chromatography – mass spectrometry; GPT: gamma-glutamyltransferase; HOMA-IR: homeostatic model assessment – insulin resistance; IR: insulin resistance; LCFA: long-chain fatty acids; LC-MS: liquid chromatography – mass spectrometry; NMR: nuclear magnetic resonance; NOB: non obese; NW: normal weight; OB: obese; OW: overweight; Phe: phenylalanine; Pro: proline; Pyr: pyruvate; SM: sphingomyelin; T2D: type 2 diabetes; TAG: triacylglycerides; Trp: tryptophan; Tyr: tyrosine.