Post-glucose Metabolite Signatures Reflect Insulin Sensitivity and Beta-cell Function in Black South African Women

Post-glucose Metabolite Signatures Reflect Insulin Sensitivity and Beta-cell Function in Black South African Women | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Post-glucose Metabolite Signatures Reflect Insulin Sensitivity and Beta-cell Function in Black South African Women Asanda Mtintsilana, Lebogang Moshupya, Yingxu Zeng, Melony Fortuin-De-Smidt, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9160380/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 9 You are reading this latest preprint version Abstract Background Postprandial metabolic responses are strong predictors of type 2 diabetes (T2D) and its underlying pathophysiological traits, insulin sensitivity and β-cell function, but remain poorly characterised in African populations. We investigated glucose-stimulated metabolite profiles across the glycaemic spectrum and their associations with insulin sensitivity and β-cell function in Black South African women. Methods This cross-sectional study included 65 women (median age: 56 years) from the Middle-Aged Soweto Cohort, categorised as normal glucose tolerance (NGT, n = 29), impaired glucose tolerance (IGT, n = 24) and T2D (n = 12). Following an oral glucose tolerance test, insulin sensitivity (Matsuda index) and β-cell function were estimated from the Mari model. Changes in metabolic profiles were characterised from fasting to 30 (Δ30) and 120 (Δ120) minutes post-glucose ingestion using a multi-platform metabolomics approach. Results Insulin sensitivity and β-cell function declined progressively from NGT to IGT to T2D. From Δ30 and Δ120, carbohydrates and bile acids increased, whereas amino acids (including BCAAs), fatty acids and lysophospholipids decreased across all groups; and associated with insulin sensitivity and β-cell function. At Δ120, bile acids, cholic acid and deoxycholic acid, remained elevated in the T2D group only. Lysophospholipids decreased across groups. Carbohydrates, bile acids, and lysophospholipids correlated positively with insulin sensitivity and β-cell function, while amino acids, BCAAs, and fatty acids correlated negatively. Conclusion: Distinct post-glucose metabolite responses across the glycaemic spectrum reflect differences in insulin sensitivity and β-cell function in Black South African women, highlighting the value of dynamic metabolic profiling for understanding T2D progression in African populations. Type 2 diabetes insulin sensitivity beta-cell function metabolomics Black women South Africa Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Type 2 diabetes (T2D) represents a significant public health challenge in South Africa, with a rising prevalence that disproportionately affects women [1, 2]. In 2020, diabetes was the leading natural cause of death among South African women [3]. Notably, the pathogenesis of T2D varies across ethnic groups [4]. For example, Black African populations tend to exhibit a phenotype characterised by low insulin sensitivity accompanied by hyperinsulinaemia, attributed to elevated insulin secretion and reduced hepatic insulin clearance[5–7]. Emerging evidence suggests that insulin hypersecretion, rather than classical peripheral insulin resistance, may be a primary initiating event in the T2D development in Black Africans[5, 8–10]. However, the mechanisms driving such early hyperinsulinaemia remain poorly understood, particularly in Black South African women, who bear a disproportionately high T2D burden and its associated risk factors [11]. Metabolomics offers a powerful lens to examine the biochemical perturbations underlying T2D risk [12–15]. Extensive research in European and Asian cohorts has established fasting metabolite biomarkers as predictors of insulin resistance and future diabetes risk [16–18]; however, such data remain sparse in African populations [19, 20]. For instance, in European and Asian populations, elevated branched-chain amino acids (BCAAs) have been linked to the development of T2D via effects on both the regulation of insulin secretion and insulin action [21]. Likewise, lysophospholipids (e.g., lysophosphatidylcholine, LPC) have been implicated in the regulation of glucose homeostasis [22], with evidence that they can induce insulin resistance through inflammatory kinase pathways [23] while also modulating insulin secretion via G-protein-coupled receptors on pancreatic β-cells [24]. Consistent with findings in other cohorts [17, 25, 26], the few studies conducted in African populations have primarily focused on the fasted state [19, 20]. These investigations identified key fasting metabolites such as BCAAs, aromatic amino acids, hexoses, bile acids, and phospholipids that are associated with prediabetes and T2D risk [19, 20]. Despite the clinical importance of post-prandial metabolism, the dynamic metabolic response to a glucose challenge remains largely underexplored[27], particularly in African populations. Consequently, there is a significant knowledge gap regarding how dynamic metabolite profiles relate to the underlying drivers of T2D pathophysiology, specifically insulin sensitivity and β-cell function. Therefore, the present study aimed to investigate metabolite profiles during an oral glucose tolerance test (OGTT) across the glycaemic spectrum and their associations with insulin sensitivity and β-cell function in middle-aged Black South African women. Materials and Methods Study design and participant selection This cross-sectional study was conducted between 2015 and 2016 and included a subgroup of 83 participants with comprehensive metabolomics data [19] from the Middle-Aged Soweto Cohort (MASC) [28]. Participants were eligible if they had stored fasting samples, were HIV seronegative and had dual-energy X-ray absorptiometry (DXA)-derived body composition measures, as previously described [19, 29]. Participants with known T2D diagnosis (n=15) and incomplete OGTT data (n=3) were excluded, resulting in a final analytical sample of 65 participants: NGT (n=29), IGT (n=24) and T2D (n=12). This sample allowed us to examine metabolite changes across the glycaemic spectrum and their associations with insulin sensitivity and β-cell function, key features of T2D pathophysiology. Socio-demographic and lifestyle factors Socio-demographic data were collected using a standard questionnaire and included measures of socio-economic status such as educational attainment, marital status and employment status. Current smoking status and alcohol consumption were also measured. Physical activity and sedentary behaviours were objectively measured using ActiGraph GT3X-Plus triaxial accelerometers (ActiGraph GT3X-Plus, ActiGraph LLC, Pensacola, Florida) and activPAL devices (activPAL3c, PAL Technologies Ltd, Glasgow, Uk) as previously described [30]. Body composition and body fat distribution Weight, height, waist, and hip circumference were measured using standard procedures, as previously described[19]. Body mass index (BMI) was calculated (weight (kg)/height (m 2 )) and classified according to World Health Organization criteria: normal weight (18.5–24.9 kg/m 2 ), overweight (25–29.9 kg/m 2 ) or obesity (≥30 kg/m 2 ) [31]. Furthermore, whole body composition was assessed using DXA with a Hologic Discovery-W (S/N 71201) system (software version 13.4.2:7). This included fat mass (FM) measured in kilograms and as a percentage of total body mass, excluding the head (sub-total) to minimise potential artifacts affecting the readings. Additionally, trunk and leg FM were measured, expressed in kilograms and as a percentage of subtotal FM. Abdominal visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) were also estimated from DXA, as previously described [32]. Blood sampling and biochemical analyses Blood samples were drawn after an overnight fast (>10 hours), following which participants completed a 75g standard OGTT, in which blood samples were collected at 10, 20, 30, 60, 90 and 120 minutes after glucose ingestion for the determination of glucose, insulin, C-peptide concentrations. Metabolomics analyses were performed on samples collected at 0, 30 and 120 minutes (as detailed below), to capture metabolite changes during the early and overall response to glucose ingestion. Plasma glucose concentrations were analysed on the Randox RX Daytona Chemistry Analyser using enzymatic methods (Randox Laboratories Ltd., London, UK). Serum insulin and C-peptide assays were analysed on the Immulite® 1000 Immunoassay System (Siemens Chemiluminescent Healthcare GmbH, Henkestr, Germany). The WHO diagnostic criteria were used for the categorisation of glycaemic groups[33] : NGT, fasting glucose < 6.1 mmol/l and/or 2-h post glucose load < 7.8 mmol/l, impaired fasting glucose (IFG) (fasting glucose 6.1-6.9 mmol/L), IGT (2-h post glucose load 7.8-11.0 mmol/l) and T2D (fasting glucose ≥ 7.0 mmol/l and/or 2-h post glucose load ≥ 11.1 mmol/l). Insulin sensitivity and β-cell function calculations The mathematical modelling of C-peptide and glucose concentrations during the OGTT was used to assess β-cell function according to Mari [34]. This model describes the relationship between insulin secretion rates (ISR) and β-cell glucose sensitivity (β-CGS) as the sum of two components [34–36]. Glucose concentration drives insulin secretion through: (1) a dose-response component, which reflects the static relationship between glucose levels and ISR, modulated by a potentiation factor which accounts for time-dependent influences such as incretins; and (2) a dynamic component, which captures the early phase insulin response triggered by rapid increases in glucose [34, 37]. From this function, β-CGS (the slope) and ISR at fixed reference glucose levels (i.e. 5 mmol/L for NGT and 7 mmol/L for T2D) were determined. Additionally, total ISR, representing the integral of ISR over the entire test, was calculated. Insulin clearance was estimated as the ratios of basal ISR to basal insulin levels and as the mean ISR/mean insulin levels during the test. Insulin sensitivity was evaluated using the Matsuda index [38]. Metabolomics analyses The sample preparation and metabolomics analysis protocols have been described in detail previously [39]. Briefly, extracted fasting and OGTT-derived serum metabolites were subjected to multi-platform metabolomics analyses including gas chromatography time-of-flight mass spectrometry (GC-TOF/MS) and liquid chromatography time-of-flight mass spectrometry (LC-TOF/MS), operating in positive and negative modes. Quality control samples (QC, pooled from all samples) and a QC-dilution series were included in each analytical batch to monitor the instrument stability, exclude background features and to assist in the identification of metabolites. Given the known methodological biases, all samples were prepared and analysed in a specific order with internal randomisation [40]. Metabolites were annotated by matching the resolved mass spectra (MS/MS when available), and retention indices to the in-house mass spectral library at the Swedish Metabolomics Centre (www.swedishmetabolomicscentre.se) and the publicly available Max Planck Institute library (http://csbdb.mpimp-golm.mpg.de/csbdb/gmd/ gmd.html) and METLIN database (http://metlin.scripps.edu/), Human Metabolome Database (HMDB) (http://www.hmdb.ca/). All LC-TOF/MS data processing was performed using MassHunter Profinder (version B.08.00), Qualitative Analysis (version B.07.00), PCDL manager (version B.07.00) and Mass Profiler Professional™ 13.0, all from Agilent Technologies Inc., Santa Clara, CA, USA. The combined use of GC-TOF/MS and LC-TOF/MS allowed us to comprehensively analyse serum metabolites with diverse chemical properties. After filtering out noisy peaks, we quantified 840 putative metabolites, of which 252 were annotated and categorised into amino acids, carbohydrates, fatty acids, lipids, acylcarnitines and organic acids, following the classification criteria of the Human Metabolome Database (www.HMDB.ca). Statistical analysis Clinical data were analysed using STATA/SE version 18. A Shapiro-Wilk test was used to assess the normality of continuous variables, with normally distributed presented as mean ± standard deviation (SD), while non-normally distributed continuous variables were reported as median (25th–75th percentiles). Categorical variables were expressed as frequencies (n) and percentages (%). One-way analysis of variance (ANOVA) was used to compare normally distributed continuous variables between different glycaemic groups, followed by Bonferroni post hoc analysis, while the Kruskal-Wallis test was used to compare non-normally distributed continuous variables, followed by Dunn's test for pairwise comparisons. The chi-squared (χ 2 ) and Fisher’s exact test, where appropriate, were used to compare categorical variables between different glycaemic groups. A significant level of p<0.05 was used for statistical significance. Metabolomics To analyse the metabolomics data, we employed a multi-step approach integrating exploratory data analysis and advanced statistical modelling. First, principal component analysis (PCA) was used for data inspection and quality control, identifying outliers and low-quality variables to ensure dataset integrity by excluding poorly reproducible metabolites. Second, we analysed the baseline (fasted) circulating metabolome using orthogonal partial least squares (OPLS) modelling to compare individuals with NGT and T2D, while the IGT group was predicted within this model to assess distinctions or overlaps in metabolite signatures. Third, we applied OPLS-effect projection (OPLS-EP) [40] to evaluate the metabolic response to an OGTT, focusing on changes from the fasted state to 30- and 120-minutes post-glucose ingestion, and assessed associations with β-cell function and insulin sensitivity. Due to the cost of non-targeted metabolomics, we focused on Δ30 minutes to capture the early post-glucose metabolic response, characterised by insulin-mediated suppression of fatty acids and amino acids metabolism and a shift towards glucose utilisation, and on Δ120 minutes to reflect the integrated metabolic response during the entire OGTT. These models included individuals from NGT, IGT and T2D groups, with sensitivity analyses excluding T2D individuals to validate robustness. Because metabolite features are highly correlated, we used projection-based modelling (OPLS) to handle multicollinearity and to improve interpretability by separating predictive variation from orthogonal (non-predictive) variation, without inflating degrees of freedom during variable selection. Model significance was assessed using CV-ANOVA, and metabolites were retained in the reported signatures only if they contributed to the predictive component (VIP >1) and their 95% jack-knifed confidence intervals did not include zero. Univariate p-values was only calculated post hoc only for filtered set of metabolites and are provided to support interpretation of the multivariate signatures. Thus, no global FDR correction across all 840 detected features was performed since metabolomics data contain strong dependency structures and non-unique features (adducts/isotopologues). Statistical analyses were conducted using MATLAB R2016a (The MathWorks, Natick, MA, USA) and SIMCA v.18 software (Sartorius Stedim, Umetrics, Sweden), with stringent filtering to ensure only reliable and biologically relevant metabolites were included in the final signatures. Results Participant characteristics according to glycaemic groups Table 1 summarises participant anthropometry, DXA, glucose and insulin by glycaemic status. There were no differences in age, weight, WC and BMI between groups. Leg fat and trunk fat (%FM) were lower, and VAT area was higher in the T2D group compared to the NGT group, while other fat measures did not differ between glycaemic groups. As expected, fasting and 2-hour glucose levels were highest in the T2D group, followed by the IGT group and then the NGT group. Fasting C-peptide and insulin levels were also higher in the T2D and IGT groups than in the NGT group (Table 1). Fig. 1 illustrates the differences in plasma glucose (A), serum insulin (B) and serum C-peptide (C) responses during an OGTT and the corresponding modelled insulin dynamics (insulin secretion rate (D) and ISR dose response (E)) across the three glycaemic groups. Between-group comparisons are summarised in Table 1: Basal ISR was highest in the T2D group, followed by the IGT and NGT groups. However, insulin secretion at both reference glucose levels (5 mmol/L and 7 mmol/L) was significantly higher in the NGT and IGT groups compared to the T2D group. Total ISR was elevated in the IGT group relative to NGT but did not differ from T2D. Insulin sensitivity and β-CGS were highest in NGT, intermediate in IGT, and lowest in T2D. Basal and total insulin clearance did not differ between groups. Socio-demographic and lifestyle characteristics are presented and compared by glycaemic group in Supplementary Table 1. There were no differences in sociodemographic factors, smoking and alcohol intake, physical activity and sedentary behaviour between the glycaemic groups. Table 1 Characteristics of participants by glycaemic categories ( n = 65) NGT IGT T2D p-value N N=29 N=24 N=12 Age (years) 56 (50-60) 56 (49-59) 58 (53-61) 0.485 Anthropometry Weight (kg) 81.7 ± 18.6 86.6 ± 15.8 93.1 ± 24.7 0.202 BMI (kg/m 2 ) 33.1 ± 6.6 34.6 ± 6.7 37.5 ± 7.7 0.187 Waist circumference (cm) 97.2 ± 15.4 101.6 ± 11.4 105.8 ± 12.6 0.166 Hip circumference (cm) 119.4 ± 13.9 120.0 ± 12.1 124.6 ± 14.4 0.509 WHR 0.81 (0.07) 0.85 (0.06) 0.85 (0.05) 0.086 BMI categories, n (%) Normal weight 5 (17.2%) 1 (4.2%) 0 (0.0%) 0.261 Overweight 4 (13.8%) 5 (20.8%) 1 (8.3%) Obese 20 (69.0%) 18 (75.0%) 11 (91.7%) DXA-derived body fat and fat distribution Body fat mass (kg) 37.3 ± 12.8 40.6 ± 11.2 44.6 ± 16.9 0.261 Body fat (%) 51.1 (43.8-53.6) 51.2 (46.6-53.2) 52.2 (45.0-55.1) 0.706 Trunk (%FM) 45.4 ± 7.0 ‡ 47.5 ± 3.5 50.7 ± 5.8 0.031 Leg (%FM) 43.3 ± 6.9 ‡ 39.9 ± 3.5 37.3 ± 5.7 0.007 VAT (cm 2 ) 151 ± 78 ‡ 174 ± 51 217± 65 0.017 SAT (cm 2 ) 483 ± 181 504 ± 140 525 ± 133 0.730 Glucose and insulin measures Fasting glucose (mmol/L) 4.7 (4.0-5.1) †, ‡ 5.4 (4.6-5.8) § 6.8 (5.7-7.4) <0.001 Fasting C-peptide (pmol/L) 559.4 (347.6-705.0) †, ‡ 849.0 (625.6-959.9) 902.0 (609.0-1570.6) 0.002 Fasting insulin (pmol/L) 58.8 (17.4-95.8) †, ‡ 92.7 (49.0-142.7) 113.2 (87.2-205.6) 0.013 Basal ISR (pmol/min/m 2 ) 75.1 (46.5-85.7) †, ‡ 108.8 (77.5-119.4) 116.9 (73.1-172.1) 0.002 ISR @ 5mmol (pmol/min/m 2 ) 152.7 (103.2-269.6) †, ‡ 108.4 (54.2-148.0) 78.6 (23.3-153.9) 0.008 ISR @7mmol (pmol/min/m 2 ) 438.4 (326.8-778.5) †, ‡ 303.8 (237.7-354.0) § 147.6 (86.6-246.6) 0.0001 Basal insulin clearance (L/ min /m 2 ) 1.2 (0.9-2.0) 1.2 (0.8-1.6) 0.9 (0.7-1.1) 0.129 2-hour glucose (mmol/L) 5.8 (4.9-7.2) †, ‡ 9.0 (8.5-9.6) § 12.2 (11.6-13.2) <0.001 Total ISR (pmol/min/m 2 ) 46.9 (38.3-67.3) † 61.3 (50.8-81.2) 53.7 (47.1-67.1) 0.040 Total insulin clearance (L/ min/m 2 ) 0.8 (0.7-1.1) 0.9 (0.8-1.1) 0.7 (0.6-0.9) 0.231 β-cell glucose sensitivity (pmol/min/m/mM) 151.1 (97.9-239.7) †, ‡ 105.7 (75.2-139.3) § 55.4 (46.1-76.1) <0.001 Insulin sensitivity (Matsuda index) 4.9 (2.5-7.5) †, ‡ 2.3 (1.9-4.2) § 1.7 (1.0-2.1) <0.001 Data presented as means ± SD or median (25 th -75 th percentile) and n (%). P-value derived from ANOVA or Kruskal–Wallis, significant post hoc differences: †NGT compared to IGT; ‡NGT compared to T2D; §IGT compared to T2D with P ≤ 0.05 deemed as statistically significant. Abbreviations: BMI, Body mass index; FM, Fat mass; HC, Hip circumference; IGT, Impaired metabolism; NGT , Normal glucose tolerance; SAT, Subcutaneous adipose tissue; T2D, Type 2 diabetes; VAT, Visceral adipose tissue; WC; Waist circumference; WHR, waist-to-hip ratio. Post-glucose ingestion metabolite responses A comprehensive overview of all significantly altered metabolites for each glycaemic group from a fasted state to 30 and 120 minutes (Δ30 and Δ120) following an OGTT is provided in Fig. 2a and 2b, respectively. Overall, the OGTT induced significant alterations in metabolite signatures that were observed across all glycaemic groups at both time points (OPLS, CV-ANOVA P<0.0001). At Δ30-minutes, increases in carbohydrate metabolites and citric acid cycle intermediates (e.g., citric acid and isocitric acid) were observed primarily in the NGT and IGT groups, whereas only selected carbohydrate metabolites (1,5-anhydroglucitol and saccharic acid) increased across all groups (Fig. 2a). At Δ120 minutes, mannitol and cis-inositol remained elevated in the NGT and IGT groups, whereas lactulose and xylitol were the only carbohydrate metabolites increased in the T2D group. However, glucosamine increased, while arabinose decreased significantly across all groups. Citric acid cycle intermediates, which increased at Δ30 minutes in NGT and IGT, were significantly reduced across all groups by Δ120 minutes (Fig. 2b). At Δ30 minutes, several amino acids and derivatives in the T2D and IGT groups were significantly reduced, changes not observed in the NGT group, except for significant increases in betaine and creatinine. By Δ120 minutes, amino acids decreased across all groups, with the largest reductions observed in the NGT group. Fatty acids, lipids, and most acylcarnitines decreased at Δ30 minutes across all groups and remained suppressed at Δ120 minutes, although at Δ30 minutes C5-carnitine and C5-iso-carnitine increased only in the NGT group. Lysophospholipids showed a significant decrease across groups at Δ30 minutes, with more pronounced reductions in the NGT and IGT groups (Fig. 2a). Notably, their precursors, glycerol-2-phosphate and glycerol-3-phosphate, were significantly reduced only in the T2D group at Δ120 minutes (Fig. 2b). Bile acids increased across all groups at both time points; however, at Δ120 minutes, only cholic acid and deoxycholic acid remained elevated in the T2D group. Supplementary Fig. 1 illustrates differences in fasting metabolite signatures across the three glycaemic groups (Supplementary Fig. 1a), variations in fasted circulating metabolites between the NGT and T2D groups (Supplementary Fig. 1b) and identifies significant metabolite profiles that differentiated NGT and T2D groups (Supplementary Fig. 1c). Unlike post-glucose responses, fasting metabolite profiles showed minimal differences between the glycaemic groups. Compared to the NGT group, individuals with T2D presented with higher BCAA metabolites, saturated fatty acids, carbohydrates, and lower lysophospholipids (Supplementary Fig. 1a). In contrast, there was no significant difference in bile acids between NGT and T2D groups. The IGT group exhibited an intermediary metabolite signature. Fig. 3 and Fig. 4 show associations between OGTT-derived metabolic responses at Δ30 and Δ120 minutes and measures of insulin sensitivity and β-cell function. OPLS correlation analyses at both time points (Fig. 3a-d; Fig. 4a-d) showed amplified metabolite responses, allowing for the detection of metabolite signatures associated with insulin sensitivity and β-cell function (OPLS CV-ANOVA, P = 0.001). We observed overlapping associations between metabolite signatures and insulin sensitivity and β-cell function, which varied at the Δ30 and Δ120 minutes post-glucose ingestion. At Δ30 and Δ120 minutes, four to five carbohydrate metabolites, including mannitol, cis-inositol, and glucosamine were positively associated with both measures of insulin sensitivity and β-cell function. However, at Δ120 minutes, only glucosamine remained positively associated with β-cell function. At Δ30-minute, only L-Methionine was negatively associated with β-cell function, whereas multiple amino acids, including L-Proline, L-Methionine, and Hippuric acid were also associated with lower insulin sensitivity. In contrast, L-kynurenine and L-tryptophan were the only amino acids positively associated with insulin sensitivity (Fig. 3a-d). Similar to the Δ30-minute findings, several amino acids, including L-Proline and L-Methionine remained negatively associated with both measures of insulin sensitivity and β-cell function at Δ120 minutes (Fig. 4a-d). Furthermore, leucine was the only BCAA negatively associated with insulin sensitivity at Δ30 minutes; however, at Δ120 minutes, multiple BCAAs and intermediates (e.g., Leucine, valine and fructosyl-leucine), were negatively associated with both insulin sensitivity and β-cell function. Long-chain acylcarnitines and fatty acids were negatively associated with insulin sensitivity and β-cell function at both time points (Fig. 3 and Fig. 4). In contrast, at the Δ30-minute time point several lysophospholipids were positively correlated with β-cell function, but not with insulin sensitivity (Fig. 3a-d). However, at Δ120 minute mark, only lysophosphatidylethanolamine (LPE (16:0)) and platelet-activating factor (PAF (18:1)) showed positive correlations with β-cell function and insulin sensitivity, respectively (Fig. 4a-d). Notably, bile acids were consistently and positively associated with both measures at both time points (Fig. 3 and Fig. 4). Discussion Our findings demonstrate that dynamic metabolite responses to a glucose challenge revealed aspects of insulin regulation. Notably, we observed subtle differences in fasting metabolite profiles between glycaemic groups, whereas glucose ingestion resulted in pronounced differences between groups. At 30 and 120 minutes post-OGTT, we identified distinct metabolite signatures and subsequently showed that these metabolite profiles were significantly associated with insulin sensitivity and β-cell function, confirming that metabolite dynamics reflect insulin secretion and action, providing a more informative view of insulin regulation in this high-risk cohort than static measures alone. Bile acids stood out as a consistent and novel signal of insulin regulation. While fasting bile acids did not differ between groups, glucose ingestion resulted in differential increases, specifically in normoglycaemic individuals at 30 minutes. By 120 minutes, conjugated bile acids (e.g. cholic and deoxycholic acid) were strikingly elevated in the T2D group relative to others, a pattern not seen at the fasted state. This exaggerated bile acid response aligns with previous findings showing elevated bile acids, particularly deoxycholic acid, in individuals with diabetes [17], while others such as 7-ketolithocholic acid and isolithocholic acid have been inversely associated with incident T2D risk [17]. Mechanistically, bile acids engage receptors such as Farnesoid X receptor (FXR) and TGR5 which regulate hepatic glucose production, peripheral insulin sensitivity and glucagon-like peptide-1 (GLP-1) [41]. Furthermore, recent evidence indicates that both fasting and postprandial bile acid profiles are closely linked to key aspects of glucose metabolism, including insulin sensitivity, incretin hormone activity and T2D risk, underscoring the importance dynamic measurements rather than static concentrations alone [42, 43]. The positive correlations we observed between post-load bile acid elevations and both insulin sensitivity and β-cell function suggest a protectivecompensatory role to maintain glucose homeostasis, particularly in those with impaired glucose metabolism. In individuals with higher insulin sensitivity and preserved β-cell function, elevated circulating bile acid levels after glucose ingestion may enhance insulin secretion via incretin-mediated [44] or direct islet effects [45] and concurrently improve peripheral insulin action. Together these mechanisms could buffer postprandial excursions. The T2D group’s sustained high levels of cholic acid and deoxycholic acid at two hours could reflect a physiological attempt to compensate for insulin resistance or an alteration in gut-liver signalling in diabetes. However, because fasting bile acids were similar across groups, these effects may be limited to the postprandial state. Differences between our findings and those of Ibrahim et al [17], likely reflect the distinction between fasting bile acids as indicators of chronic metabolic exposure and post-glucose bile acid responses as markers of acute glucose-stimulated signalling related to insulin sensitivity and β-cell function. Overall, bile acid dynamics appear to encode information about overall insulin responsiveness, with greater post-load increases associated with better insulin sensitivity and secretion, highlighting their potential role as modulators or biomarkers of insulin efficacy. BCAA and related metabolites were linked with lower insulin sensitivity and lower β-cell function, suggesting metabolic stress on the insulin-producing cells. Fasting BCAA levels were modestly higher in the T2D group, consistent with prior associations between BCAAs and diabetes risk [19, 46, 47]. After the glucose load, BCAA levels (and their keto-acid derivatives) were reduced across all groups, reflecting insulin-mediated uptake and suppression of proteolysis, but this reduction was significantly blunted in the T2D group. By 120 minutes, women with NGT showed the largest BCAA reductions, whereas those with T2D showed the smallest, corroborating earlier findings that impaired glucose tolerance and T2D associate with a diminished BCAA clearance following nutrient challenges [48–50]. Evidence of efficient BCAA catabolism, reflected by a transient rise of valine/isoleucine-derived C5 acylcarnitines at 30 minutes was observed only in the NGT group, suggesting that metabolically healthy individuals readily channel BCAAs into oxidative pathways. In contrast, its absence in IGT/T2D implies a block or saturation in BCAA catabolic flux, which has been linked to insulin resistance progression [46, 51]. At 30 minutes, leucine levels were selectively and negatively associated with insulin sensitivity (but not with β-cell function), whereas by 120 minutes, multiple BCAAs (leucine, isoleucine, valine and their intermediates) were inversely correlated with both insulin sensitivity and β-cell function. This suggests that leucine elevated in the early postprandial state may specifically flag insulin resistance, even before broad β-cell stress manifests, in line with leucine’s known capacity to impair insulin signalling in muscle [52]. By the later postprandial phase, i.e., 120min, persistent elevations of BCAA metabolites may signal a more global metabolic dysregulation affecting both insulin action and secretion. We suggest that BCAA dynamics indicate overall insulin “strain” on the system: when insulin action was effective (NGT), BCAAs were efficiently cleared; when insulin action was impaired or β-cells overwhelmed (IGT/T2D), BCAA levels remained inappropriately high. These results support the concept that BCAAs are not just bystanders but can actively contribute to insulin resistance and β-cell dysfunction [46, 53, 54]. The observed positive associations between carbohydrate metabolites and tricarboxylic acid (TCA) cycle intermediates with insulin dynamics, alongside the negative associations between fatty acids and amino acids with insulin dynamics, particularly at 30-minutes post-glucose ingestion, are consistent with the physiological role of insulin action in promoting glucose utilisation while suppressing lipolysis and proteolysis during the early postprandial period. Broad OGTT-induced changes across metabolic pathways, including carbohydrate and TCA cycle metabolites, have been consistently reported in metabolic profiling studies [47]. In participants with NGT, this was reflected by robust increases in carbohydrate and related metabolites, whereas these responses were attenuated in individuals with T2D, suggesting impaired postprandial glucose handling and downstream oxidative metabolism, potentially linked to β-cell dysfunction and insulin resistance. Accordingly, lower levels of 1,5-anhydroglucitol in individuals with T2D are consistent with findings from a Nigerian population, where this metabolite was significantly reduced in T2D [20]. Given that 1,5-anhydroglucitol is a marker of short-term glycaemic control and postprandial hyperglycaemia [55], its reduction in T2D further supports the presence of dysregulated glucose metabolism in this group. Moreover, the exclusive elevation of glucosamine in T2D at Δ120 suggests a selective preservation of glucosamine responsiveness in T2D, highlighting its potential role in glucose sensing and β-cell function. Correlation analyses further supported this, showing that several carbohydrate metabolites were positively associated with measures of insulin dynamics at both Δ30, whereas at Δ120, only glucosamine remained linked to β-cell function. Together with evidence from other populations showing that glucosamine supplementation is associated with lower T2D risk [56, 57], these results suggest that endogenous glucosamine dynamic may reflect a compensatory glucose-sensing pathway that remains active in T2D despite broader impairments in postprandial carbohydrate metabolism. Lysophospholipids exhibited a distinct pattern linked to insulin regulation. We observed that several lysophosphatidylcholines and lysophosphatidylethanolamines (LPCs and LPEs) decreased after glucose ingestion in the NGT and IGT groups, whereas the T2D group only showed subtle changes in lysophospholipid levels. Instead, the T2D group showed a reduction in certain precursors like glycerol-3-phosphate. This response differs from that reported in some healthy European-ancestry populations, where postprandial LPC levels tend to rise in response to a glucose challenge [58]. The drop in lysophospholipids among the Black South African women with NGT/IGT may indicate a healthy metabolic flexibility or tissue uptake of these lipids in response to insulin. Prior work in this cohort noted that lower fasting LPC levels were associated with future T2D development [19], indicating that inadequate lysophospholipid availability could drive compensatory hypersecretion of insulin [59]. Consistent with their dual roles, we found that at 30 min post-load, higher levels of several lysophospholipids correlated with higher β-cell function, suggesting that lysophospholipids may acutely support insulin secretion, possibly by serving as signalling molecules or fuel substrates for β-cells. However, by 120 min, only a couple of lipid mediators remained significantly associated with insulin dynamics: for example, LPE (16:0) was positively associated with β-cell function, and platelet-activating factor (PAF 18:1, a specialised phospholipid) was positively associated with insulin sensitivity. These time-dependent associations suggest that lysophospholipid metabolism is linked to both insulin secretion and insulin sensitivity. Higher levels of certain lysophospholipids may support the early postprandial β-cell response, whereas later, other lipids appear more related to insulin action in peripheral tissues. The generally blunted lysophospholipid response in T2D (compared with NGT/IGT) indicates disturbed lipid handling, which may both reflect and promote β-cell dysfunction and insulin resistance in this group, highlighting these lipids as potential early biomarkers of T2D in this population. Collectively, the post-OGTT metabolite responses, notably bile acids, BCAA-derivatives, carbohydrates and lysophospholipids, provide an integrated description of variations in β-cell function and insulin sensitivity. Unlike fasting metabolomics, this dynamic approach captures perturbations specifically related to glucose handling, where defects in insulin action or release manifest as altered metabolite excursions that are not evident at baseline. For example, the attenuated suppression of fatty acids and BCAAs in T2D reflects impaired insulin action despite similar fasting levels, underscoring the added value of post-challenge profiling. Thus, quantifying the magnitude and direction of changes in metabolites such as BCAAs or bile acids after glucose loading could help distinguish predominant defects in insulin sensitivity versus β-cell function, with potential implications for personalised interventions. Moreover, associations between specific metabolite responses (e.g. bile acids) and more favourable insulin sensitivity and secretion highlight candidate pathways for therapeutic modulation. Several limitations should be noted. While the study provides novel insights into dynamic metabolic response to glucose ingestion across glycaemic groups, the exploratory nature of non-targeted metabolomics and the relatively small sample size, highlight the need for future studies to replicate and extend these analyses in larger, independent cohorts and to evaluate their potential utility as biomarkers of T2D. Similar to other large sample size studies such as those by Liu et al., 2023 [27], future studies should consider multiple sampling across the OGTT to further characterise the temporal dynamics of metabolite responses in African populations. Additionally, while the study focuses on Black South African women, the sample may not represent the full ethnic and genetic diversity of Black African women, potentially limiting generalisability. Finally, the cross-sectional design restricts the ability to track metabolic changes over time, particularly in relation to T2D progression. The population context is indeed important. Black South African women often present with hyperinsulinaemia [59], and the distinct post-glucose lipid and amino acid profiles observed here may reflect population-specific adaptive or maladaptive mechanisms. These data emphasise that T2D pathophysiology cannot be fully captured by uniform models and that population-specific metabolic profiling is essential. By linking dynamic metabolite signatures to insulin sensitivity and secretion in this high-risk group, our study identifies potential early markers of T2D risk and motivates longitudinal and interventional studies to test their predictive and mechanistic relevance. Conclusions This study highlights the value of dynamic metabolomic profiling during an OGTT in uncovering early metabolic disturbances linked to insulin resistance and β-cell dysfunction in Black SA women. Shifts in bile acids, lysophospholipids, BCCAs, fatty acids and acylcarnitines following an OGTT revealed distinct metabolic signatures associated with glycaemic status and insulin dynamics, observations that are not apparent in the fasting state. These findings suggest that specific metabolite responses to glucose may serve as early biomarkers of T2D progression and development and offer potential targets for intervention. Future studies should focus on longitudinal designs to monitor the changes in metabolic signatures over time and their role in T2D development and progression. Abbreviations β-cell, Beta-cell; β-CGS, Beta-cell glucose sensitivity; BCAAs, Branched-chain amino acids; BMI, Body mass index; DXA, Dual-energy X-ray absorptiometry; FM, Fat mass; FXR, Farnesoid X receptor; GC-TOF/MS, Gas chromatography time-of-flight mass spectrometry; HC, Hip circumference; IFG, Impaired fasting glucose; IGT, Impaired glucose tolerance; ISR, Insulin secretion rates; JNK, c-Jun N-terminal kinase; LC-TOF/MS, Liquid chromatography time-of-flight mass spectrometry; LPC, Lysophosphatidylcholine; LPE, Lysophosphatidylethanolamine; MASC, Middle-Aged Soweto Cohort; NGT, Normal glucose tolerance; OGTT, Oral glucose tolerance test; OPLS, Orthogonal partial least squares; OPLS-EP, OPLS-effect Projection; PAF, Platelet-activating factor; PCA, Principal component analysis; PE, Phosphatidylethanolamine; SAT, Subcutaneous adipose tissue; TGR, Takeda G protein-coupled receptor 5; VAT, Visceral adipose tissue; VIP, Variable importance in projection; WC, Waist circumference; WHR, Waist-to-hip ratio Declarations Acknowledgements We would like to express our sincere gratitude to all the women in the study for their invaluable participation and contribution to this study. We also extend our appreciation to the research team for their unwavering dedication, and the Swedish Metabolomics Centre (http://www.swedishmetabolomicscentre.se/) is acknowledged for access to instrumentation and technical support. We are also grateful to the South African Medical Research Council (SAMRC)/University of the Witwatersrand (WITS) Developmental Pathways for Health Research Unit (DPHRU) at the Chris Hani Baragwanath Hospital in Soweto, Johannesburg, SA, for providing the resources and support necessary for the successful completion of this work. Author contribution statement E.C., J.H.G., L.K.M., and T.O. contributed to the conception and design of the study. Y.Z. and E.C. carried out metabolomics data analysis and the development of the corresponding methodology. L.M. performed the clinical data analysis and drafted the initial version of the manuscript. A.M. was responsible for data collection, sample processing, and clinical data analysis, and critically reviewed and edited the manuscript. K.M.U. conducted the Mari Modelling. J.H.G., L.K.M., M.F, E.C, Y.Z, K.M.U and T.O. provided critical review and revisions of the manuscript. E.C. is the guarantor of this work, with full access to all the data included in the study and assumes responsibility for the integrity and accuracy of the data and analyses. Statement of Ethics The study was approved by the Human Research Ethics Committee (Medical) of the University of the Witwatersrand (M150530). The procedures and risks associated with the study were explained to the participants and they all provided signed informed consent prior to participation in the study. All testing procedures were performed at the South African Medical Research Council (SAMRC)/University of the Witwatersrand Developmental Pathways for Health Research Unit (DPHRU) at the Chris Hani Baragwanath Hospital in Soweto Johannesburg. Funding This research was supported by the Swedish Research Council (Swedish Development Grant, DNR: 2014-2522) and the National Research Foundation (NRF) of SA, which provided a scholarship to Asanda Mtintsilana (grant number 111308). Data availability Original data generated and analysed during this study are available in a public repository. All processed metabolomics data, sample metadata, and metabolomics annotations have been deposited in Zenodo (doi:10.5281/zenodo.16977693). During peer review, access is provided to editors and reviewers via a private link; upon publication the repository will be released publicly under CC BY 4.0. Statements and Declarations The authors have no competing interests to declare that are relevant to the content of this article. References Pheiffer C, Wyk VP Van, Turawa E, Levitt N, Kengne AP, Bradshaw D (2021) Prevalence of type 2 diabetes in South Africa: A systematic review and meta-analysis. Int J Environ Res Public Health 18(11):5868. https://doi.org/10.3390/IJERPH18115868/S1 International Diabetes Federation. IDF Atlas 11th Edition (2025) IDF Diabetes Atlas, 11th Edition. https://diabetesatlas.org/media/uploads/sites/3/2025/04/IDF_Atlas_11th_Edition_2025-1.pdf. Accessed 27 Jan 2026 Statistics South Africa (2020) STATISTICAL RELEASE Mortality and causes of death in South Africa: Findings from death notification 2020. www.statssa.gov.za, [email protected] ,Tel+27123108911 Hasson BR, Apovian C, Istfan N (2015) Racial/Ethnic Differences in Insulin Resistance and Beta Cell Function: Relationship to Racial Disparities in Type 2 Diabetes among African Americans versus Caucasians. Curr Obes Rep 4(2):241–249. https://doi.org/10.1007/S13679-015-0150-2/METRICS Goedecke JH, Olsson T (2020) Pathogenesis of type 2 diabetes risk in black Africans: a South African perspective. J Intern Med 288(3):284–294. https://doi.org/10.1111/JOIM.13083 Goedecke JH, Mendham AE (2022) Pathophysiology of type 2 diabetes in sub-Saharan Africans. Diabetologia 65(12):1967. https://doi.org/10.1007/S00125-022-05795-2 Ladwa M, Bello O, Hakim O, et al (2020) Insulin clearance as the major player in the hyperinsulinaemia of black African men without diabetes. Diabetes Obes Metab 22(10):1808–1817. https://doi.org/10.1111/DOM.14101 Esser N, Utzschneider KM, Kahn SE (2020) Early beta cell dysfunction vs insulin hypersecretion as the primary event in the pathogenesis of dysglycaemia. Diabetologia 63(10):2007–2021. https://doi.org/10.1007/s00125-020-05245-x Tricò D, Natali A, Arslanian S, Mari A, Ferrannini E (2018) Identification, pathophysiology, and clinical implications of primary insulin hypersecretion in nondiabetic adults and adolescents. JCI Insight 3(24). https://doi.org/10.1172/jci.insight.124912 Bergman RN, Piccinini F, Kabir M, Kolka CM, Ader M (2019) Hypothesis: Role of Reduced Hepatic Insulin Clearance in the Pathogenesis of Type 2 Diabetes. Diabetes 68(9):1709–1716. https://doi.org/10.2337/DB19-0098 Goedecke JH, Hellig J, Conradie-Smit M, May W (2025) Epidemiology of adult obesity. S Afr Med J 115(9b):16–25. https://doi.org/10.7196/SAMJ.2025.v115i9b.3616 Shi L, Brunius C, Lehtonen M, et al (2018) Plasma metabolites associated with type 2 diabetes in a Swedish population: a case–control study nested in a prospective cohort. Diabetologia 61(4):849–861. https://doi.org/10.1007/S00125-017-4521-Y/FIG.S/3 Ahola-Olli A V., Mustelin L, Kalimeri M, et al (2019) Circulating metabolites and the risk of type 2 diabetes: a prospective study of 11,896 young adults from four Finnish cohorts. Diabetologia 62(12):2298–2309. https://doi.org/10.1007/S00125-019-05001-W/TABLES/2 Urpi-Sarda M, Almanza-Aguilera E, Llorach R, et al (2019) Non-targeted metabolomic biomarkers and metabotypes of type 2 diabetes: A cross-sectional study of PREDIMED trial participants. Diabetes Metab 45(2):167–174. https://doi.org/10.1016/J.DIABET.2018.02.006 Lu Y, Wang Y, Ong CN, et al (2016) Metabolic signatures and risk of type 2 diabetes in a Chinese population: an untargeted metabolomics study using both LC-MS and GC-MS. Diabetologia 59(11):2349–2359. https://doi.org/10.1007/S00125-016-4069-2/FIG.S/4 Morze J, Wittenbecher C, Schwingshackl L, et al (2022) Metabolomics and Type 2 Diabetes Risk: An Updated Systematic Review and Meta-analysis of Prospective Cohort Studies. Diabetes Care 45(4):1013–1024. https://doi.org/10.2337/dc21-1705 Choucair I, Mallela DP, Hilser JR, et al (2024) Comprehensive Clinical and Genetic Analyses of Circulating Bile Acids and Their Associations With Diabetes and Its Indices. Diabetes 73(8):1215–1228. https://doi.org/10.2337/DB23-0676 Jiang X, Zhu F, Graça G, et al (2025) Serum Metabolomic Profiling of Incident Type 2 Diabetes Mellitus in the Multi-ethnic Study of Atherosclerosis and Rotterdam Study. J Clin Endocrinol Metab 110(8):e2700–e2710. https://doi.org/10.1210/CLINEM/DGAE812 Zeng Y, Mtintsilana A, Goedecke JH, Micklesfield LK, Olsson T, Chorell E (2019) 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 95:57–64. https://doi.org/10.1016/j.metabol.2019.04.001 Doumatey AP, Shriner D, Zhou J, et al (2024) Untargeted metabolomic profiling reveals molecular signatures associated with type 2 diabetes in Nigerians. Genome Med 16(1):1–18. https://doi.org/10.1186/S13073-024-01308-5/FIG.S/7 Yang J, Chi Y, Burkhardt BR, Guan Y, Wolf BA (2010) Leucine metabolism in regulation of insulin secretion from pancreatic beta cells. Nutr Rev 68(5):270–279. https://doi.org/10.1111/J.1753-4887.2010.00282.X Yea K, Kim J, Yoon JH, et al (2009) Lysophosphatidylcholine activates adipocyte glucose uptake and lowers blood glucose levels in murine models of diabetes. Journal of Biological Chemistry 284(49):33833–33840. https://doi.org/10.1074/jbc.M109.024869 Han MS, Lim YM, Quan W, et al (2011) Lysophosphatidylcholine as an effector of fatty acid-induced insulin resistance. J Lipid Res 52(6):1234–1246. https://doi.org/10.1194/JLR.M014787 Drzazga A, Kristinsson H, Sałaga M, et al (2018) Lysophosphatidylcholine and its phosphorothioate analogues potentiate insulin secretion via GPR40 (FFAR1), GPR55 and GPR119 receptors in a different manner. Mol Cell Endocrinol 472:117–125. https://doi.org/10.1016/J.MCE.2017.12.002 Wang X-H, Xu F, Cheng M, et al Fasting serum total bile acid levels are associated with insulin sensitivity, islet β-cell function and glucagon levels in response to glucose challenge in patients with. jstage.jst.go.jpX Wang, F Xu, M Cheng, X Wang, D Zhang, L Zhao, H Cai, H Huang, T Chen, X ZhangEndocrine Journal, 2020•jstage.jst.go.jp 2020(11):1107–1117 Guasch-Ferré M, Hruby A, Toledo E, et al (2016) Metabolomics in Prediabetes and Diabetes: A Systematic Review and Meta-analysis. Diabetes Care 39(5):833–846. https://doi.org/10.2337/DC15-2251 Liu Y, Liu X, He Q, Huang X, Ren Y, Dong Z (2024) Changes in Isoleucine, Sarcosine, and Dimethylglycine During OGTT as Risk Factors for Diabetes. J Clin Endocrinol Metab 109(7):1793–1802. https://doi.org/10.1210/CLINEM/DGAE018 Goedecke JH, Nguyen KA, Kufe C, et al (2022) Waist circumference thresholds predicting incident dysglycaemia and type 2 diabetes in Black African men and women. Diabetes Obes Metab 24(5):918–927. https://doi.org/10.1111/dom.14655 Mtintsilana A, Micklesfield LK, Chorell E, Olsson T, Goedecke JH (2019) Fat redistribution and accumulation of visceral adipose tissue predicts type 2 diabetes risk in middle-aged black South African women: a 13-year longitudinal study. Nutr Diabetes 9(1):12. https://doi.org/10.1038/S41387-019-0079-8 Micklesfield LK, Westgate K, Smith A, et al (2022) Physical Activity Behaviors of a Middle-Age South African Cohort as Determined by Integrated Hip and Thigh Accelerometry. Med Sci Sports Exerc 54(9):1493–1505. https://doi.org/10.1249/MSS.0000000000002940 World Health Organization (2000) Obesity:preventing and managing the global epidemic. Geneva, Switzerland Micklesfield LK, Goedecke JH, Punyanitya M, Wilson KE, Kelly TL (2012) Dual-Energy X-Ray Performs as Well as Clinical Computed Tomography for the Measurement of Visceral Fat. Obesity 20(5):1109–1114. https://doi.org/10.1038/OBY.2011.367 Balkau B, Charles MA (1999) Comment on the provisional report from the WHO consultation. Diabetic Medicine 16(5):442–443. https://doi.org/10.1046/j.1464-5491.1999.00059.x Mari A, Ferrannini E (2008) β-cell function assessment from modelling of oral tests: an effective approach. Diabetes Obes Metab 10(SUPPL. 4):77–87. https://doi.org/10.1111/J.1463-1326.2008.00946.X Mari A, Schmitz O, Gastaldelli A, Oestergaard T, Nyholm B, Ferrannini E (2002) Meal and oral glucose tests for assessment of beta -cell function: modeling analysis in normal subjects. Am J Physiol Endocrinol Metab 283(6). https://doi.org/10.1152/ajpendo.00093.2002 Van Cauter E, Mestrez F, Sturis J, Polonsky KS (1992) Estimation of insulin secretion rates from C-peptide levels. Comparison of individual and standard kinetic parameters for C-peptide clearance. Diabetes 41(3):368–377. https://doi.org/10.2337/DIAB.41.3.368 Mari A, Tura A, Grespan E, Bizzotto R (2020) Mathematical Modeling for the Physiological and Clinical Investigation of Glucose Homeostasis and Diabetes. Front Physiol 11:575789. https://doi.org/10.3389/FPHYS.2020.575789/XML/NLM Matsuda M, DeFronzo RA (1999) Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22(9):1462–1470. https://doi.org/10.2337/DIACARE.22.9.1462 Jiye A, Trygg J, Gullberg J, et al (2005) Extraction and GC/MS analysis of the human blood plasma metabolome. Anal Chem 77(24):8086–8094. https://doi.org/10.1021/ac051211v Jonsson P, Wuolikainen A, Thysell E, et al (2015) Constrained randomization and multivariate effect projections improve information extraction and biomarker pattern discovery in metabolomics studies involving dependent samples. Metabolomics 11(6):1667–1678. https://doi.org/10.1007/S11306-015-0818-3/FIG.S/5 Kim H, Fang S (2018) Crosstalk between FXR and TGR5 controls glucagon-like peptide 1 secretion to maintain glycemic homeostasis. Lab Anim Res 34(4):140–146. https://doi.org/10.5625/lar.2018.34.4.140 Hansen M, Sonne DP, Mikkelsen KH, Gluud LL, Vilsbøll T, Knop FK (2017) Bile acid sequestrants for glycemic control in patients with type 2 diabetes: A systematic review with meta-analysis of randomized controlled trials. J Diabetes Complications 31(5):918–927. https://doi.org/10.1016/J.JDIACOMP.2017.01.011 Cadena Sandoval M, Haeusler RA (2025) Bile acid metabolism in type 2 diabetes mellitus. Nat Rev Endocrinol 21(4):203–213. https://doi.org/10.1038/S41574-024-01067-8;SUBJMETA Ticho AL, Malhotra P, Dudeja PK, Gill RK, Alrefai WA (2019) Bile acid receptors and gastrointestinal functions. Liver Res 3(1):31–39. https://doi.org/10.1016/j.livres.2019.01.001 Düfer M, Hörth K, Wagner R, et al (2012) Bile acids acutely stimulate insulin secretion of mouse β-cells via farnesoid X receptor activation and K(ATP) channel inhibition. Diabetes 61(6):1479–1489. https://doi.org/10.2337/DB11-0815 Newgard CB, An J, Bain JR, et al (2009) A Branched-Chain Amino Acid-Related Metabolic Signature that Differentiates Obese and Lean Humans and Contributes to Insulin Resistance. Cell Metab 9(4):311–326. https://doi.org/10.1016/J.CMET.2009.02.002/ATTACHMENT/2556051D-632E-4829-BEF5-65236D548452/MMC1.PDF Ho JE, Larson MG, Vasan RS, et al (2013) Metabolite profiles during oral glucose challenge. Diabetes 62(8):2689–2698. https://doi.org/10.2337/DB12-0754/-/DC1 Sjögren RJO, Rizo-Roca D, Chibalin A V., et al (2021) Branched-chain amino acid metabolism is regulated by ERRα in primary human myotubes and is further impaired by glucose loading in type 2 diabetes. Diabetologia 64(9):2077–2091. https://doi.org/10.1007/S00125-021-05481-9 Fanni G, Eriksson JW, Pereira MJ (2023) Several Metabolite Families Display Inflexibility during Glucose Challenge in Patients with Type 2 Diabetes: An Untargeted Metabolomics Study. Metabolites 13(1):131. https://doi.org/10.3390/METABO13010131/S1 Gannon NP, Schnuck JK, Vaughan RA (2018) BCAA Metabolism and Insulin Sensitivity – Dysregulated by Metabolic Status? Mol Nutr Food Res 62(6):1700756. https://doi.org/10.1002/MNFR.201700756 Adams SH (2011) Emerging Perspectives on Essential Amino Acid Metabolism in Obesity and the Insulin-Resistant State. Advances in Nutrition 2(6):445. https://doi.org/10.3945/AN.111.000737 Krebs M, Krssak M, Bernroider E, et al (2002) Mechanism of amino acid-induced skeletal muscle insulin resistance in humans. Diabetes 51(3):599–605. https://doi.org/10.2337/DIABETES.51.3.599 Ferrannini E, Natali A, Camastra S, et al (2013) Early metabolic markers of the development of dysglycemia and type 2 diabetes and their physiological significance. Diabetes 62(5):1730–1737. https://doi.org/10.2337/DB12-0707/-/DC1 Lynch CJ, Adams SH (2014) Branched-chain amino acids in metabolic signalling and insulin resistance. Nat Rev Endocrinol 10(12):723. https://doi.org/10.1038/NRENDO.2014.171 Kim MJ, Jung HS, Hwang-Bo Y, et al (2013) Evaluation of 1,5-anhydroglucitol as a marker for glycemic variability in patients with type 2 diabetes mellitus. Acta Diabetol 50(4):505–510. https://doi.org/10.1007/S00592-011-0302-0 Ma H, Li X, Zhou T, et al (2020) Glucosamine Use, Inflammation, and Genetic Susceptibility, and Incidence of Type 2 Diabetes: A Prospective Study in UK Biobank. Diabetes Care 43(4):719–725. https://doi.org/10.2337/DC19-1836 Zhou S, Zhou P, Yang T, Si J, An W, Jiang Y (2025) Glucosamine supplementation contributes to reducing the risk of type 2 diabetes: Evidence from Mendelian randomization combined with a meta-analysis. J Int Med Res 53(4):03000605251334460. https://doi.org/10.1177/03000605251334460 Zhao X, Peter A, Fritsche J, et al (2009) Changes of the plasma metabolome during an oral glucose tolerance test: Is there more than glucose to look at? Am J Physiol Endocrinol Metab 296(2):384–393. https://doi.org/10.1152/AJPENDO.90748.2008/ASSET/IMAGES/LARGE/ZH10020955660008.JPEG Goedecke JH, Dave JA, Faulenbach M V., et al (2009) Insulin response in relation to insulin sensitivity: an appropriate beta-cell response in black South African women. Diabetes Care 32(5):860–865. https://doi.org/10.2337/dc08-2048 Additional Declarations No competing interests reported. Supplementary Files STROBEchecklist.docx SUPPLEMENTARY.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 01 May, 2026 Reviews received at journal 21 Apr, 2026 Reviews received at journal 19 Apr, 2026 Reviewers agreed at journal 14 Apr, 2026 Reviewers agreed at journal 14 Apr, 2026 Reviewers invited by journal 01 Apr, 2026 Editor assigned by journal 31 Mar, 2026 Submission checks completed at journal 31 Mar, 2026 First submitted to journal 25 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9160380","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":616728595,"identity":"fe821599-0e72-4057-be8a-56f480b3b305","order_by":0,"name":"Asanda Mtintsilana","email":"","orcid":"","institution":"South African Medical Research Council","correspondingAuthor":false,"prefix":"","firstName":"Asanda","middleName":"","lastName":"Mtintsilana","suffix":""},{"id":616728596,"identity":"a6c4c1f1-f653-4a35-a0ce-76e83f515bb7","order_by":1,"name":"Lebogang Moshupya","email":"","orcid":"","institution":"University of the Witwatersrand","correspondingAuthor":false,"prefix":"","firstName":"Lebogang","middleName":"","lastName":"Moshupya","suffix":""},{"id":616728597,"identity":"c338ac26-8ee7-4a5b-be60-c56df352ffc6","order_by":2,"name":"Yingxu Zeng","email":"","orcid":"","institution":"Hainan Tropical Ocean University","correspondingAuthor":false,"prefix":"","firstName":"Yingxu","middleName":"","lastName":"Zeng","suffix":""},{"id":616728598,"identity":"a2239531-6498-4e53-881f-a761d9de87cc","order_by":3,"name":"Melony Fortuin-De-Smidt","email":"","orcid":"","institution":"Umeå University","correspondingAuthor":false,"prefix":"","firstName":"Melony","middleName":"","lastName":"Fortuin-De-Smidt","suffix":""},{"id":616728599,"identity":"76609674-f747-46a0-ba24-40c8860d826d","order_by":4,"name":"Lisa K. Micklesfield","email":"","orcid":"","institution":"University of the Witwatersrand","correspondingAuthor":false,"prefix":"","firstName":"Lisa","middleName":"K.","lastName":"Micklesfield","suffix":""},{"id":616728600,"identity":"aadefb45-d78e-47e1-8f81-192134623c3a","order_by":5,"name":"Tommy Olsson","email":"","orcid":"","institution":"Umeå University","correspondingAuthor":false,"prefix":"","firstName":"Tommy","middleName":"","lastName":"Olsson","suffix":""},{"id":616728601,"identity":"660ed34a-cf7e-4d81-b070-629cf4dc9fa5","order_by":6,"name":"Kristina M. Utzschneider","email":"","orcid":"","institution":"VA Puget Sound Health Care System","correspondingAuthor":false,"prefix":"","firstName":"Kristina","middleName":"M.","lastName":"Utzschneider","suffix":""},{"id":616728602,"identity":"da0b7f1e-b613-453b-9dee-8e54782fc032","order_by":7,"name":"Julia H. Goedecke","email":"","orcid":"","institution":"University of the Witwatersrand","correspondingAuthor":false,"prefix":"","firstName":"Julia","middleName":"H.","lastName":"Goedecke","suffix":""},{"id":616728603,"identity":"e8e533a2-6866-4c56-a3a8-1c514d0feaed","order_by":8,"name":"Elin Chorell","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0UlEQVRIiWNgGAWjYDACdsYGIHkAiJkbD3wgSgszXAtjw8EZxGkBkxAth3mI0cHPzNz2gKHmjpx5e2PDYdu2bYkN/IcP4NUi2czYbsBw7JmxzJmDDYdz224nNkikJeDVYnCYsU0C6KTEGRKJMC08Bni12MO1yD9sOGwJ0sJ/Hn/AGTDDbQGSjCAtDDl4dTBIgGxJOHbYWIInseFgz7nbxm0Safgdxt/e/kziQ81hOQn2wwcf/Ci7LdvPf/gBfmtAIAGZw0ZY/SgYBaNgFIwCQgAAp/5KuzZD5N0AAAAASUVORK5CYII=","orcid":"","institution":"Umeå University","correspondingAuthor":true,"prefix":"","firstName":"Elin","middleName":"","lastName":"Chorell","suffix":""}],"badges":[],"createdAt":"2026-03-18 14:10:57","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9160380/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9160380/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106340136,"identity":"ce9a484e-aaeb-4446-b7f8-bfaed8b2dc54","added_by":"auto","created_at":"2026-04-07 15:27:41","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":247367,"visible":true,"origin":"","legend":"\u003cp\u003ePlasma glucose (A), serum insulin (B), serum C-peptide (C), insulin secretion rate (D) and ISR glucose dose response (E) during an oral glucose tolerance test in participants with normal glucose tolerance (NGT), impaired glucose metabolism (IGT) and type 2 diabetes (T2D). Beta-cell glucose sensitivity (β-CGS) is the mean slope of the ISR glucose dose-response curve. Data are mean ± SEM.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9160380/v1/9f71047b31d3f3071fe35ba5.png"},{"id":106340107,"identity":"aed85273-eec8-4255-8e38-0dc410e7e7d3","added_by":"auto","created_at":"2026-04-07 15:27:35","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":455362,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea \u003c/strong\u003eHeat map showing significantly altered metabolites highlighted from an OPLS model that calculated the response in metabolites from a fasted state to \u003cem\u003e30 minutes after an oral glucose tolerance\u003c/em\u003e test in individuals with type 2 diabetes (T2D), impaired glucose tolerance (IGT) and normal glucose tolerance (NGT). Metabolites are grouped by class, and the values are represented as -log10(P). White indicates no statistical significance (p \u0026lt; 0.05), while red and blue represent significantly higher and lower levels of metabolites, respectively. Light and dark shades indicate the least and most significant changes, respectively.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb \u003c/strong\u003eHeat map showing significantly altered metabolites highlighted from an OPLS model that calculated the response in metabolites from a fasted state to \u003cem\u003e120 minutes after an oral glucose tolerance\u003c/em\u003e test in individuals with type 2 diabetes (T2D), impaired glucose tolerance (IGT) and normal glucose tolerance (NGT). Metabolites are grouped by class, and the values are represented as -log10(P). White indicates no statistical significance (p \u0026lt; 0.05), while red and blue represent significantly higher and lower levels of metabolites, respectively. Light and dark shades indicate the least and most significant changs, respectively\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9160380/v1/6351eb20728f410b7d8dc0c3.png"},{"id":106340118,"identity":"ef84b8f0-e214-46df-964e-e33c75078112","added_by":"auto","created_at":"2026-04-07 15:27:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":425792,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea-d\u003c/strong\u003e Associations between the Δ30-minute metabolome response to an OGTT and physiological estimates of beta-cell function and insulin sensitivity. (a, c) OPLS models illustrating associations between the Δ30-minute metabolome response and beta-cell function (a) and insulin sensitivity (c), calculated across all glycaemic groups. The Y-axis represents variable importance in projection (VIP) values, while the X-axis displays OPLS weights (w*). Metabolites significantly associated with each physiological measure are highlighted, as determined by jack-knifing confidence intervals at the 95% significance level. (b, d). Detailed metabolome response signatures for beta-cell function (b) and insulin sensitivity (d), showing metabolites with significant association. A positive w* suggest a positive association with beta-cell function (b) or insulin sensitivity (d), whereas a negative w* suggests a negative association between the metabolites and insulin dynamics. Metabolites are categorised by biochemical class: amino acids, branched-chain amino acids (BCAA), carbohydrates, acylcarnitines, fatty acids, lysophospholipids, bile acids, and unclassified metabolites. \u003cem\u003eKey abbreviations: OPLS, orthogonal partial least squares; PC, phosphatidylcholine; IAA, isoleucine; BCAA, branched-chain amino acids; G-DCA, glycodeoxycholic acid; T-DCA, taurodeoxycholic acid; T-CDCA, taurochenodeoxycholic acid; G-CDCA, glycochenodeoxycholic acid; G-CA, glycocholic acid; 1,5-AG, 1,5-anhydroglucitol.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9160380/v1/95e9abf6449562576cca4119.png"},{"id":106340106,"identity":"75497a8e-2276-43e8-a22e-fbcc9e6dae44","added_by":"auto","created_at":"2026-04-07 15:27:35","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":542736,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea-d\u003c/strong\u003e Associations between the Δ120-minute metabolome response to an OGTT and physiological estimates of beta-cell function and insulin sensitivity. OPLS models illustrating associations between the Δ120-minute metabolome response and beta-cell function (a) and insulin sensitivity (c). The Y-axis represents variable importance in projection (VIP) values, while the X-axis displays OPLS weights (w*). Metabolites significantly associated with each physiological measure are highlighted, as determined by jack-knifing confidence intervals at the 95% significance level. (b, d). Detailed metabolome response signatures for beta-cell function (b) and insulin sensitivity (d), showing metabolites with significant associations. A positive w* suggest a positive association with beta-cell function (b) or insulin sensitivity (d), whereas a negative w* suggests a negative association between the metabolites and insulin dynamics. Metabolites are categorised by biochemical class: amino acids, branched-chain amino acids (BCAA), carbohydrates, acylcarnitines, fatty acids, lysophospholipids, bile acids, and unclassified metabolites. \u003cem\u003eKey abbreviations: OPLS, orthogonal partial least squares; PC, phosphatidylcholine; BCAA, branched-chain amino acids; G-DCA, glycodeoxycholic acid; T-DCA, taurodeoxycholic acid; T-CDCA, taurochenodeoxycholic acid; G-CDCA, glycochenodeoxycholic acid; G-CA, glycocholic acid.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-9160380/v1/9e83512db5b962b27d9233a0.png"},{"id":106340223,"identity":"a207b554-ddc1-4815-9f43-c1624b7a6119","added_by":"auto","created_at":"2026-04-07 15:28:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2418842,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9160380/v1/b81794ff-cd12-46af-8221-8a235b39ab29.pdf"},{"id":106340103,"identity":"e762a179-6bf2-45b3-9d7e-1c45e0fb0325","added_by":"auto","created_at":"2026-04-07 15:27:35","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":34752,"visible":true,"origin":"","legend":"","description":"","filename":"STROBEchecklist.docx","url":"https://assets-eu.researchsquare.com/files/rs-9160380/v1/6043386b173cc7c4c6dc2e4d.docx"},{"id":106340104,"identity":"52259288-3633-4e7d-a526-9aaf8ba19edf","added_by":"auto","created_at":"2026-04-07 15:27:35","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":168680,"visible":true,"origin":"","legend":"","description":"","filename":"SUPPLEMENTARY.docx","url":"https://assets-eu.researchsquare.com/files/rs-9160380/v1/d3ab05d9c0935a452af0c5c3.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003ePost-glucose Metabolite Signatures Reflect Insulin Sensitivity and Beta-cell Function in Black South African Women\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eType 2 diabetes (T2D) represents a significant public health challenge in South Africa, with a rising prevalence that disproportionately affects women [1, 2]. In 2020, diabetes was the leading natural cause of death among South African women [3]. Notably, the pathogenesis of T2D varies across ethnic groups [4]. For example, Black African populations tend to exhibit a phenotype characterised by low insulin sensitivity accompanied by hyperinsulinaemia, attributed to elevated insulin secretion and reduced hepatic insulin clearance[5\u0026ndash;7]. Emerging evidence suggests that insulin hypersecretion, rather than classical peripheral insulin resistance, may be a primary initiating event in the T2D development in Black Africans[5, 8\u0026ndash;10]. However, the mechanisms driving such early hyperinsulinaemia remain poorly understood, particularly in Black South African women, who bear a disproportionately high T2D burden and its associated risk factors [11].\u003c/p\u003e \u003cp\u003eMetabolomics offers a powerful lens to examine the biochemical perturbations underlying T2D risk [12\u0026ndash;15]. Extensive research in European and Asian cohorts has established fasting metabolite biomarkers as predictors of insulin resistance and future diabetes risk [16\u0026ndash;18]; however, such data remain sparse in African populations [19, 20]. For instance, in European and Asian populations, elevated branched-chain amino acids (BCAAs) have been linked to the development of T2D via effects on both the regulation of insulin secretion and insulin action [21]. Likewise, lysophospholipids (e.g., lysophosphatidylcholine, LPC) have been implicated in the regulation of glucose homeostasis [22], with evidence that they can induce insulin resistance through inflammatory kinase pathways [23] while also modulating insulin secretion via G-protein-coupled receptors on pancreatic β-cells [24]. Consistent with findings in other cohorts [17, 25, 26], the few studies conducted in African populations have primarily focused on the fasted state [19, 20]. These investigations identified key fasting metabolites such as BCAAs, aromatic amino acids, hexoses, bile acids, and phospholipids that are associated with prediabetes and T2D risk [19, 20]. Despite the clinical importance of post-prandial metabolism, the dynamic metabolic response to a glucose challenge remains largely underexplored[27], particularly in African populations. Consequently, there is a significant knowledge gap regarding how dynamic metabolite profiles relate to the underlying drivers of T2D pathophysiology, specifically insulin sensitivity and β-cell function. Therefore, the present study aimed to investigate metabolite profiles during an oral glucose tolerance test (OGTT) across the glycaemic spectrum and their associations with insulin sensitivity and β-cell function in middle-aged Black South African women.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003e\u0026nbsp;Study design and participant selection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis cross-sectional study was conducted between 2015 and 2016 and included a subgroup of 83 participants with comprehensive metabolomics data [19] from the Middle-Aged Soweto Cohort (MASC) [28]. Participants were eligible if they had stored fasting samples, were HIV seronegative and had dual-energy X-ray absorptiometry (DXA)-derived body composition measures, as previously described [19, 29]. Participants with known T2D diagnosis (n=15) and incomplete OGTT data (n=3) were excluded, resulting in a final analytical sample of 65 participants: NGT (n=29), IGT (n=24) and T2D (n=12). This sample allowed us to examine metabolite changes across the glycaemic spectrum and their associations with insulin sensitivity and \u0026beta;-cell function, key features of T2D pathophysiology.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSocio-demographic and lifestyle factors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSocio-demographic data were collected using a standard questionnaire and included measures of socio-economic status such as educational attainment, marital status and employment status. Current smoking status and alcohol consumption were also measured. \u0026nbsp;Physical activity and sedentary behaviours were objectively measured using ActiGraph GT3X-Plus triaxial accelerometers (ActiGraph GT3X-Plus, ActiGraph LLC, Pensacola, Florida) and activPAL devices (activPAL3c, PAL Technologies Ltd, Glasgow, Uk) as previously described [30].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBody composition and body fat distribution\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWeight, height, waist, and hip circumference were measured using standard procedures, as previously described[19]. Body mass index (BMI) was calculated (weight (kg)/height\u003csup\u003e\u0026nbsp;\u003c/sup\u003e(m\u003csup\u003e2\u003c/sup\u003e)) and classified according to World Health Organization criteria: normal weight (18.5\u0026ndash;24.9 kg/m\u003csup\u003e2\u003c/sup\u003e), overweight (25\u0026ndash;29.9 kg/m\u003csup\u003e2\u003c/sup\u003e) or obesity (\u0026ge;30 kg/m\u003csup\u003e2\u003c/sup\u003e) [31]. Furthermore, whole body composition was assessed using DXA with a Hologic Discovery-W (S/N 71201) system (software version 13.4.2:7). This included fat mass (FM) measured in kilograms and as a percentage of total body mass, excluding the head (sub-total) to minimise potential artifacts affecting the readings. Additionally, trunk and leg FM were measured, expressed in kilograms and as a percentage of subtotal FM. Abdominal visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) were also estimated from DXA, as previously described [32]. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBlood sampling and biochemical analyses\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBlood samples were drawn after an overnight fast (\u0026gt;10 hours), following which participants completed a 75g standard OGTT, in which blood samples were collected at 10, 20, 30, 60, 90 and 120 minutes after glucose ingestion for the determination of glucose, insulin, C-peptide concentrations. Metabolomics analyses were performed on samples collected at 0, 30 and 120 minutes (as detailed below), to capture metabolite changes during the early and overall response to glucose ingestion. Plasma glucose concentrations were analysed on the Randox RX Daytona Chemistry Analyser using enzymatic methods (Randox Laboratories Ltd., London, UK). Serum insulin and C-peptide assays were analysed on the Immulite\u0026reg; 1000 Immunoassay System (Siemens Chemiluminescent Healthcare GmbH, Henkestr, Germany).\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe WHO diagnostic criteria were used for the categorisation of glycaemic groups[33] : NGT, fasting glucose \u0026lt; 6.1 mmol/l and/or 2-h post glucose load \u0026lt; 7.8 mmol/l, impaired fasting glucose (IFG) (fasting glucose 6.1-6.9 mmol/L), IGT (2-h post glucose load 7.8-11.0 mmol/l) and T2D (fasting glucose \u0026ge; 7.0 mmol/l and/or 2-h post glucose load \u0026ge; 11.1 mmol/l). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInsulin sensitivity and \u0026beta;-cell function calculations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe mathematical modelling of C-peptide and glucose concentrations during the OGTT was used to assess \u0026beta;-cell function according to Mari [34]. This model describes the relationship between insulin secretion rates (ISR) and \u0026beta;-cell glucose sensitivity (\u0026beta;-CGS) as the sum of two components [34\u0026ndash;36]. Glucose concentration drives insulin secretion through: (1) a dose-response component, which reflects the static relationship between glucose levels and ISR, modulated by a potentiation factor which accounts for time-dependent influences such as incretins; and (2) a dynamic component, which captures the early phase insulin response triggered by rapid increases in glucose [34, 37]. From this function,\u0026nbsp;\u0026beta;-CGS\u0026nbsp;(the slope) and ISR at fixed reference glucose levels (i.e. 5 mmol/L for NGT and 7 mmol/L for T2D) were determined. Additionally, total ISR, representing the integral of ISR over the entire test, was calculated. Insulin clearance was estimated as the ratios of basal ISR to basal insulin levels and as the mean ISR/mean insulin levels during the test. Insulin sensitivity was evaluated using the Matsuda index [38].\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMetabolomics analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe sample preparation and metabolomics analysis protocols have been described in detail previously [39]. Briefly, extracted fasting and OGTT-derived serum metabolites were subjected to multi-platform metabolomics analyses including gas chromatography time-of-flight mass spectrometry (GC-TOF/MS) and liquid chromatography time-of-flight mass spectrometry (LC-TOF/MS), operating in positive and negative modes. Quality control samples (QC, pooled from all samples) and a QC-dilution series were included in each analytical batch to monitor the instrument stability, exclude background features and to assist in the identification of metabolites. Given the known methodological biases, all samples were prepared and analysed in a specific order with internal randomisation [40]. Metabolites were annotated by matching the resolved mass spectra (MS/MS when available), and retention indices to the in-house mass spectral library at the Swedish Metabolomics Centre (www.swedishmetabolomicscentre.se) and the publicly available Max Planck Institute library (http://csbdb.mpimp-golm.mpg.de/csbdb/gmd/ gmd.html) and METLIN database (http://metlin.scripps.edu/), Human Metabolome Database (HMDB) (http://www.hmdb.ca/). All LC-TOF/MS data processing was performed using MassHunter Profinder (version B.08.00), Qualitative Analysis (version B.07.00), PCDL manager (version B.07.00) and Mass Profiler Professional\u0026trade; 13.0, all from Agilent Technologies Inc., Santa Clara, CA, USA. The combined use of GC-TOF/MS and LC-TOF/MS allowed us to comprehensively analyse serum metabolites with diverse chemical properties. After filtering out noisy peaks, we quantified 840 putative metabolites, of which 252 were annotated and categorised into amino acids, carbohydrates, fatty acids, lipids, acylcarnitines and organic acids, following the classification criteria of the Human Metabolome Database (www.HMDB.ca). \u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eClinical data were analysed using STATA/SE version 18. A Shapiro-Wilk test was used to assess the normality of continuous variables, with normally distributed presented as mean \u0026plusmn; standard deviation (SD), while non-normally distributed continuous variables were reported as median (25th\u0026ndash;75th percentiles). Categorical variables were expressed as frequencies (n) and percentages (%). One-way analysis of variance (ANOVA) was used to compare normally distributed continuous variables between different glycaemic groups, followed by Bonferroni post hoc analysis, while the Kruskal-Wallis test was used to compare non-normally distributed continuous variables, followed by Dunn\u0026apos;s test for pairwise comparisons. The chi-squared (\u0026chi;\u003csup\u003e2\u003c/sup\u003e) and Fisher\u0026rsquo;s exact test, where appropriate, were used to compare categorical variables between different glycaemic groups. \u0026nbsp;A significant level of p\u0026lt;0.05 was used for statistical significance. \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMetabolomics\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo analyse the metabolomics data, we employed a multi-step approach integrating exploratory data analysis and advanced statistical modelling. First, principal component analysis (PCA) was used for data inspection and quality control, identifying outliers and low-quality variables to ensure dataset integrity by excluding poorly reproducible metabolites. Second, we analysed the baseline (fasted) circulating metabolome using orthogonal partial least squares (OPLS) modelling to compare individuals with NGT and T2D, while the IGT group was predicted within this model to assess distinctions or overlaps in metabolite signatures. Third, we applied OPLS-effect projection (OPLS-EP) [40] to evaluate the metabolic response to an OGTT, focusing on changes from the fasted state to 30- and 120-minutes post-glucose ingestion, and assessed associations with \u0026beta;-cell function and insulin sensitivity. Due to the cost of non-targeted metabolomics, we focused on \u0026Delta;30 minutes to capture the early post-glucose metabolic response, characterised by insulin-mediated suppression of fatty acids and amino acids metabolism and a shift towards glucose utilisation, and on \u0026Delta;120 minutes to reflect the integrated metabolic response during the entire OGTT. These models included individuals from NGT, IGT and T2D groups, with sensitivity analyses excluding T2D individuals to validate robustness. Because metabolite features are highly correlated, we used projection-based modelling (OPLS) to handle multicollinearity and to improve interpretability by separating predictive variation from orthogonal (non-predictive) variation, without inflating degrees of freedom during variable selection. Model significance was assessed using CV-ANOVA, and metabolites were retained in the reported signatures only if they contributed to the predictive component (VIP \u0026gt;1) and their 95% jack-knifed confidence intervals did not include zero. Univariate p-values was only calculated post hoc only for filtered set of metabolites and are provided to support interpretation of the multivariate signatures. Thus, no global FDR correction across all 840 detected features was performed since metabolomics data contain strong dependency structures and non-unique features (adducts/isotopologues). Statistical analyses were conducted using MATLAB R2016a (The MathWorks, Natick, MA, USA) and SIMCA v.18 software (Sartorius Stedim, Umetrics, Sweden), with stringent filtering to ensure only reliable and biologically relevant metabolites were included in the final signatures.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eParticipant characteristics according to glycaemic groups\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTable 1 summarises participant anthropometry, DXA, glucose and insulin by glycaemic status. There were no differences in age, weight, WC and BMI between groups. Leg fat and trunk fat (%FM) were lower, and VAT area was higher in the T2D group compared to the NGT group, while other fat measures did not differ between glycaemic groups. As expected, fasting and 2-hour glucose levels were highest in the T2D group, followed by the IGT group and then the NGT group. Fasting C-peptide and insulin levels were also higher in the T2D and IGT groups than in the NGT group (Table 1).\u003c/p\u003e\n\u003cp\u003eFig. 1 illustrates the differences in plasma glucose (A), serum insulin (B) and serum C-peptide (C) responses during an OGTT and the corresponding modelled insulin dynamics (insulin secretion rate (D) and ISR dose response (E))\u0026nbsp;across the three glycaemic groups. Between-group comparisons are summarised in Table 1: Basal ISR was highest in the T2D group, followed by the IGT and NGT groups. However, insulin secretion at both reference glucose levels (5 mmol/L and 7 mmol/L) was significantly higher in the NGT and IGT groups compared to the T2D group. Total ISR was elevated in the IGT group relative to NGT but did not differ from T2D. Insulin sensitivity and \u0026beta;-CGS were highest in NGT, intermediate in IGT, and lowest in T2D. Basal and total insulin clearance did not differ between groups.\u003c/p\u003e\n\u003cp\u003eSocio-demographic and lifestyle characteristics are presented and compared by glycaemic group in Supplementary Table 1. There were no differences in sociodemographic factors, smoking and alcohol intake, physical activity and sedentary behaviour between the glycaemic groups.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1\u003c/strong\u003e Characteristics of participants by glycaemic categories (\u003cem\u003en\u0026nbsp;\u003c/em\u003e= 65)\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNGT\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eIGT\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eT2D\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ep-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eN=29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eN=24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eN=12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eAge (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e56 (50-60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e56 (49-59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e58 (53-61)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.485\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\" valign=\"top\" style=\"width: 601px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAnthropometry\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eWeight (kg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e81.7 \u0026plusmn; 18.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e86.6 \u0026plusmn; 15.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e93.1 \u0026plusmn; 24.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.202\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eBMI (kg/m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e33.1 \u0026plusmn; 6.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e34.6 \u0026plusmn; 6.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e37.5 \u0026plusmn; 7.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.187\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eWaist circumference (cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e97.2 \u0026plusmn; 15.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e101.6 \u0026plusmn; 11.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e105.8 \u0026plusmn; 12.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.166\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eHip circumference (cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e119.4 \u0026plusmn; 13.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e120.0 \u0026plusmn; 12.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e124.6 \u0026plusmn; 14.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.509\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eWHR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.81 (0.07)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e0.85 (0.06)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.85 (0.05)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.086\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\" valign=\"top\" style=\"width: 601px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBMI categories, n (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eNormal weight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e5 (17.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e1 (4.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0 (0.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"3\" valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e0.261\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eOverweight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e4 (13.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e5 (20.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1 (8.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eObese\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e20 (69.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e18 (75.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e11 (91.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\" valign=\"top\" style=\"width: 601px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDXA-derived body fat and fat distribution\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eBody fat mass (kg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e37.3 \u0026plusmn; 12.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e40.6 \u0026plusmn; 11.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e44.6 \u0026plusmn; 16.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.261\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eBody fat (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e51.1\u003c/p\u003e\n \u003cp\u003e(43.8-53.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e51.2\u003c/p\u003e\n \u003cp\u003e(46.6-53.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e52.2\u003c/p\u003e\n \u003cp\u003e(45.0-55.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.706\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eTrunk (%FM)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e45.4 \u0026plusmn; 7.0\u003csup\u003e\u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e47.5 \u0026plusmn; 3.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e50.7 \u0026plusmn; 5.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.031\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eLeg (%FM)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e43.3 \u0026plusmn; 6.9\u003csup\u003e\u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e39.9 \u0026plusmn; 3.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e37.3 \u0026plusmn; 5.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.007\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eVAT (cm\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e151 \u0026plusmn; 78\u003csup\u003e\u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e174 \u0026plusmn; 51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e217\u0026plusmn; 65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.017\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eSAT (cm\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e483 \u0026plusmn; 181\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e504 \u0026plusmn; 140\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e525 \u0026plusmn; 133\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.730\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"5\" valign=\"top\" style=\"width: 601px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eGlucose and insulin measures\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eFasting glucose\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;(mmol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e4.7\u003c/p\u003e\n \u003cp\u003e(4.0-5.1)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e5.4\u003c/p\u003e\n \u003cp\u003e(4.6-5.8)\u003csup\u003e\u0026nbsp;\u0026sect;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e6.8\u003c/p\u003e\n \u003cp\u003e(5.7-7.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eFasting C-peptide (pmol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e559.4\u003c/p\u003e\n \u003cp\u003e(347.6-705.0)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e849.0\u003c/p\u003e\n \u003cp\u003e(625.6-959.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e902.0\u003c/p\u003e\n \u003cp\u003e(609.0-1570.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eFasting insulin\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;(pmol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e58.8\u003c/p\u003e\n \u003cp\u003e(17.4-95.8)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e92.7\u003c/p\u003e\n \u003cp\u003e(49.0-142.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e113.2\u003c/p\u003e\n \u003cp\u003e(87.2-205.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eBasal ISR\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(pmol/min/m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e75.1\u003c/p\u003e\n \u003cp\u003e(46.5-85.7)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e108.8\u003c/p\u003e\n \u003cp\u003e(77.5-119.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e116.9\u003c/p\u003e\n \u003cp\u003e(73.1-172.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eISR @ 5mmol\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(pmol/min/m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e152.7\u003c/p\u003e\n \u003cp\u003e(103.2-269.6)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e108.4\u003c/p\u003e\n \u003cp\u003e(54.2-148.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e78.6\u003c/p\u003e\n \u003cp\u003e(23.3-153.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eISR @7mmol\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;(pmol/min/m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e438.4\u003c/p\u003e\n \u003cp\u003e(326.8-778.5)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e303.8\u003c/p\u003e\n \u003cp\u003e(237.7-354.0)\u003csup\u003e\u0026nbsp;\u0026sect;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e147.6\u003c/p\u003e\n \u003cp\u003e(86.6-246.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eBasal insulin clearance\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(L/ min\u003csup\u003e\u0026nbsp;\u003c/sup\u003e/m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.2\u003c/p\u003e\n \u003cp\u003e(0.9-2.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e1.2\u003c/p\u003e\n \u003cp\u003e(0.8-1.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.9\u003c/p\u003e\n \u003cp\u003e(0.7-1.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.129\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003e2-hour glucose\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;(mmol/L) \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e5.8\u003c/p\u003e\n \u003cp\u003e(4.9-7.2)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e9.0\u003c/p\u003e\n \u003cp\u003e(8.5-9.6)\u003csup\u003e\u0026nbsp;\u0026sect;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e12.2\u003c/p\u003e\n \u003cp\u003e(11.6-13.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eTotal ISR\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(pmol/min/m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e46.9\u003c/p\u003e\n \u003cp\u003e(38.3-67.3)\u003csup\u003e\u0026nbsp;\u0026dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e61.3\u003c/p\u003e\n \u003cp\u003e(50.8-81.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e53.7\u003c/p\u003e\n \u003cp\u003e(47.1-67.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.040\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003eTotal insulin clearance\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;(L/ min/m\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.8\u003c/p\u003e\n \u003cp\u003e(0.7-1.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e0.9\u003c/p\u003e\n \u003cp\u003e(0.8-1.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.7\u003c/p\u003e\n \u003cp\u003e(0.6-0.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e0.231\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026beta;-cell glucose sensitivity (pmol/min/m/mM)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e151.1\u003c/p\u003e\n \u003cp\u003e(97.9-239.7)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e105.7\u003c/p\u003e\n \u003cp\u003e(75.2-139.3)\u003csup\u003e\u0026nbsp;\u0026sect;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e55.4\u003c/p\u003e\n \u003cp\u003e(46.1-76.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026nbsp;Insulin sensitivity\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;(Matsuda index)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e4.9\u003c/p\u003e\n \u003cp\u003e(2.5-7.5)\u003csup\u003e\u0026nbsp;\u0026dagger;, \u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e2.3\u003c/p\u003e\n \u003cp\u003e(1.9-4.2)\u003csup\u003e\u0026nbsp;\u0026sect;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.7\u003c/p\u003e\n \u003cp\u003e(1.0-2.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 63px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eData presented as means \u0026plusmn; SD or median (25\u003csup\u003eth\u003c/sup\u003e-75\u003csup\u003eth\u003c/sup\u003e percentile) and n (%). P-value derived from ANOVA or Kruskal\u0026ndash;Wallis, significant post hoc differences: \u0026dagger;NGT compared to IGT; \u0026Dagger;NGT compared to T2D; \u0026sect;IGT compared to T2D with P \u0026le; 0.05 deemed as statistically significant.\u003cem\u003e\u0026nbsp;Abbreviations: BMI, Body mass index; FM, Fat mass; HC, Hip circumference; IGT,\u0026nbsp;\u003c/em\u003e\u003cem\u003eImpaired metabolism;\u0026nbsp;\u003c/em\u003e\u003cem\u003eNGT\u003c/em\u003e\u003cem\u003e, Normal glucose tolerance;\u003c/em\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cem\u003eSAT, Subcutaneous adipose tissue;\u0026nbsp;\u003c/em\u003e\u003cem\u003eT2D, Type 2 diabetes;\u003c/em\u003e\u003cem\u003e\u0026nbsp;VAT, Visceral adipose tissue; WC; Waist circumference; WHR, waist-to-hip ratio.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePost-glucose ingestion metabolite responses\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eA comprehensive overview of all significantly altered metabolites for each glycaemic group from a fasted state to 30 and 120 minutes (\u0026Delta;30 and \u0026Delta;120) following an OGTT is provided in Fig. 2a and 2b, respectively. Overall, the OGTT induced significant alterations in metabolite signatures that were observed across all glycaemic groups at both time points (OPLS, CV-ANOVA P\u0026lt;0.0001). At \u0026Delta;30-minutes, increases in carbohydrate metabolites and citric acid cycle intermediates (e.g., citric acid and isocitric acid) were observed primarily in the NGT and IGT groups, whereas only selected carbohydrate metabolites (1,5-anhydroglucitol and saccharic acid) increased across all groups (Fig. 2a). At \u0026Delta;120 minutes, mannitol and cis-inositol remained elevated in the NGT and IGT groups, whereas lactulose and xylitol were the only carbohydrate metabolites increased in the T2D group. However, glucosamine increased, while arabinose decreased significantly across all groups. Citric acid cycle intermediates, which increased at \u0026Delta;30 minutes in NGT and IGT, were significantly reduced across all groups by \u0026Delta;120 minutes (Fig. 2b).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAt \u0026Delta;30 minutes, several amino acids and derivatives in the T2D and IGT groups were significantly reduced, changes not observed in the NGT group, except for significant increases in betaine and creatinine. By \u0026Delta;120 minutes, amino acids decreased across all groups, with the largest reductions observed in the NGT group. Fatty acids, lipids, and most acylcarnitines decreased at \u0026Delta;30 minutes across all groups and remained suppressed at \u0026Delta;120 minutes, although at \u0026Delta;30 minutes C5-carnitine and C5-iso-carnitine increased only in the NGT group. Lysophospholipids showed a significant decrease across groups at \u0026Delta;30 minutes, with more pronounced reductions in the NGT and IGT groups (Fig. 2a). Notably, their precursors, glycerol-2-phosphate and glycerol-3-phosphate, were significantly reduced only in the T2D group at \u0026Delta;120 minutes (Fig. 2b). \u0026nbsp;Bile acids increased across all groups at both time points; however, at \u0026Delta;120 minutes, only cholic acid and deoxycholic acid remained elevated in the T2D group.\u003c/p\u003e\n\u003cp\u003eSupplementary Fig. 1 illustrates differences in fasting metabolite signatures across the three glycaemic groups (Supplementary Fig. 1a), variations in fasted circulating metabolites between the NGT and T2D groups (Supplementary Fig. 1b) and identifies significant metabolite profiles that differentiated NGT and T2D groups (Supplementary Fig. 1c). Unlike post-glucose responses, fasting metabolite profiles showed minimal differences between the glycaemic groups. Compared to the NGT group, individuals with T2D presented with higher BCAA metabolites, saturated fatty acids, carbohydrates, and lower lysophospholipids (Supplementary Fig. 1a). In contrast, there was no significant difference in bile acids between NGT and T2D groups. The IGT group exhibited an intermediary metabolite\u0026nbsp;signature.\u003c/p\u003e\n\u003cp\u003eFig. 3 and Fig. 4 show associations between OGTT-derived metabolic responses at \u0026Delta;30 and \u0026Delta;120 minutes and measures of insulin sensitivity and \u0026beta;-cell function. OPLS correlation analyses at both time points (Fig. 3a-d; Fig. 4a-d) showed amplified metabolite responses, allowing for the detection of metabolite signatures associated with insulin sensitivity and \u0026beta;-cell function (OPLS CV-ANOVA, P = 0.001). We observed overlapping associations between metabolite signatures and insulin sensitivity and \u0026beta;-cell function, which varied at the \u0026Delta;30 and \u0026Delta;120 minutes post-glucose ingestion. At \u0026Delta;30 and \u0026Delta;120 minutes, four to five carbohydrate metabolites, including mannitol, cis-inositol, and glucosamine were positively associated with both measures of insulin sensitivity and \u0026beta;-cell function. However, at \u0026Delta;120 minutes, only glucosamine remained positively associated with \u0026beta;-cell function.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAt \u0026Delta;30-minute, only L-Methionine was negatively associated with \u0026beta;-cell function, whereas multiple amino acids, including L-Proline, L-Methionine, and Hippuric acid were also associated with lower insulin sensitivity. In contrast, L-kynurenine and L-tryptophan were the only amino acids positively associated with insulin sensitivity (Fig. 3a-d). Similar to the \u0026Delta;30-minute findings, several amino acids, including L-Proline and L-Methionine remained negatively associated with both measures of insulin sensitivity and \u0026beta;-cell function at \u0026Delta;120 minutes (Fig. 4a-d). Furthermore, leucine was the only BCAA negatively associated with insulin sensitivity at \u0026Delta;30 minutes; however, at \u0026Delta;120 minutes, multiple BCAAs and intermediates (e.g., Leucine, valine and fructosyl-leucine), were negatively associated with both insulin sensitivity and \u0026beta;-cell function.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eLong-chain acylcarnitines and fatty acids were negatively associated with insulin sensitivity and \u0026beta;-cell function at both time points (Fig. 3 and Fig. 4). In contrast, at the \u0026Delta;30-minute time point several lysophospholipids were positively correlated with \u0026beta;-cell function, but not with insulin sensitivity (Fig. 3a-d). However, at \u0026Delta;120 minute mark, only lysophosphatidylethanolamine (LPE (16:0)) and platelet-activating factor (PAF (18:1)) showed positive correlations with \u0026beta;-cell function and insulin sensitivity, respectively (Fig. 4a-d). Notably, bile acids were consistently and positively associated with both measures at both time points (Fig. 3 and Fig. 4). \u003cem\u003e\u003cbr clear=\"all\"\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur findings demonstrate that dynamic metabolite responses to a glucose challenge revealed aspects of insulin regulation. Notably, we observed subtle differences in fasting metabolite profiles between glycaemic groups, whereas glucose ingestion resulted in pronounced differences between groups. At 30 and 120 minutes post-OGTT, we identified distinct metabolite signatures and subsequently showed that these metabolite profiles were significantly associated with insulin sensitivity and β-cell function, confirming that metabolite dynamics reflect insulin secretion and action, providing a more informative view of insulin regulation in this high-risk cohort than static measures alone.\u003c/p\u003e \u003cp\u003eBile acids stood out as a consistent and novel signal of insulin regulation. While fasting bile acids did not differ between groups, glucose ingestion resulted in differential increases, specifically in normoglycaemic individuals at 30 minutes. By 120 minutes, conjugated bile acids (e.g. cholic and deoxycholic acid) were strikingly elevated in the T2D group relative to others, a pattern not seen at the fasted state. This exaggerated bile acid response aligns with previous findings showing elevated bile acids, particularly deoxycholic acid, in individuals with diabetes [17], while others such as 7-ketolithocholic acid and isolithocholic acid have been inversely associated with incident T2D risk [17].\u003c/p\u003e \u003cp\u003eMechanistically, bile acids engage receptors such as Farnesoid X receptor (FXR) and TGR5 which regulate hepatic glucose production, peripheral insulin sensitivity and glucagon-like peptide-1 (GLP-1) [41]. Furthermore, recent evidence indicates that both fasting and postprandial bile acid profiles are closely linked to key aspects of glucose metabolism, including insulin sensitivity, incretin hormone activity and T2D risk, underscoring the importance dynamic measurements rather than static concentrations alone [42, 43]. The positive correlations we observed between post-load bile acid elevations and both insulin sensitivity and β-cell function suggest a protectivecompensatory role to maintain glucose homeostasis, particularly in those with impaired glucose metabolism. In individuals with higher insulin sensitivity and preserved β-cell function, elevated circulating bile acid levels after glucose ingestion may enhance insulin secretion via incretin-mediated [44] or direct islet effects [45] and concurrently improve peripheral insulin action. Together these mechanisms could buffer postprandial excursions.\u003c/p\u003e \u003cp\u003eThe T2D group\u0026rsquo;s sustained high levels of cholic acid and deoxycholic acid at two hours could reflect a physiological attempt to compensate for insulin resistance or an alteration in gut-liver signalling in diabetes. However, because fasting bile acids were similar across groups, these effects may be limited to the postprandial state. Differences between our findings and those of Ibrahim et al [17], likely reflect the distinction between fasting bile acids as indicators of chronic metabolic exposure and post-glucose bile acid responses as markers of acute glucose-stimulated signalling related to insulin sensitivity and β-cell function. Overall, bile acid dynamics appear to encode information about overall insulin responsiveness, with greater post-load increases associated with better insulin sensitivity and secretion, highlighting their potential role as modulators or biomarkers of insulin efficacy.\u003c/p\u003e \u003cp\u003eBCAA and related metabolites were linked with lower insulin sensitivity and lower β-cell function, suggesting metabolic stress on the insulin-producing cells. Fasting BCAA levels were modestly higher in the T2D group, consistent with prior associations between BCAAs and diabetes risk [19, 46, 47]. After the glucose load, BCAA levels (and their keto-acid derivatives) were reduced across all groups, reflecting insulin-mediated uptake and suppression of proteolysis, but this reduction was significantly blunted in the T2D group. By 120 minutes, women with NGT showed the largest BCAA reductions, whereas those with T2D showed the smallest, corroborating earlier findings that impaired glucose tolerance and T2D associate with a diminished BCAA clearance following nutrient challenges [48\u0026ndash;50].\u003c/p\u003e \u003cp\u003eEvidence of efficient BCAA catabolism, reflected by a transient rise of valine/isoleucine-derived C5 acylcarnitines at 30 minutes was observed only in the NGT group, suggesting that metabolically healthy individuals readily channel BCAAs into oxidative pathways. In contrast, its absence in IGT/T2D implies a block or saturation in BCAA catabolic flux, which has been linked to insulin resistance progression [46, 51]. At 30 minutes, leucine levels were selectively and negatively associated with insulin sensitivity (but not with β-cell function), whereas by 120 minutes, multiple BCAAs (leucine, isoleucine, valine and their intermediates) were inversely correlated with both insulin sensitivity and β-cell function.\u003c/p\u003e \u003cp\u003eThis suggests that leucine elevated in the early postprandial state may specifically flag insulin resistance, even before broad β-cell stress manifests, in line with leucine\u0026rsquo;s known capacity to impair insulin signalling in muscle [52]. By the later postprandial phase, i.e., 120min, persistent elevations of BCAA metabolites may signal a more global metabolic dysregulation affecting both insulin action and secretion. We suggest that BCAA dynamics indicate overall insulin \u0026ldquo;strain\u0026rdquo; on the system: when insulin action was effective (NGT), BCAAs were efficiently cleared; when insulin action was impaired or β-cells overwhelmed (IGT/T2D), BCAA levels remained inappropriately high. These results support the concept that BCAAs are not just bystanders but can actively contribute to insulin resistance and β-cell dysfunction [46, 53, 54].\u003c/p\u003e \u003cp\u003eThe observed positive associations between carbohydrate metabolites and tricarboxylic acid (TCA) cycle intermediates with insulin dynamics, alongside the negative associations between fatty acids and amino acids with insulin dynamics, particularly at 30-minutes post-glucose ingestion, are consistent with the physiological role of insulin action in promoting glucose utilisation while suppressing lipolysis and proteolysis during the early postprandial period. Broad OGTT-induced changes across metabolic pathways, including carbohydrate and TCA cycle metabolites, have been consistently reported in metabolic profiling studies [47]. In participants with NGT, this was reflected by robust increases in carbohydrate and related metabolites, whereas these responses were attenuated in individuals with T2D, suggesting impaired postprandial glucose handling and downstream oxidative metabolism, potentially linked to β-cell dysfunction and insulin resistance. Accordingly, lower levels of 1,5-anhydroglucitol in individuals with T2D are consistent with findings from a Nigerian population, where this metabolite was significantly reduced in T2D [20]. Given that 1,5-anhydroglucitol is a marker of short-term glycaemic control and postprandial hyperglycaemia [55], its reduction in T2D further supports the presence of dysregulated glucose metabolism in this group. Moreover, the exclusive elevation of glucosamine in T2D at Δ120 suggests a selective preservation of glucosamine responsiveness in T2D, highlighting its potential role in glucose sensing and β-cell function. Correlation analyses further supported this, showing that several carbohydrate metabolites were positively associated with measures of insulin dynamics at both Δ30, whereas at Δ120, only glucosamine remained linked to β-cell function. Together with evidence from other populations showing that glucosamine supplementation is associated with lower T2D risk [56, 57], these results suggest that endogenous glucosamine dynamic may reflect a compensatory glucose-sensing pathway that remains active in T2D despite broader impairments in postprandial carbohydrate metabolism.\u003c/p\u003e \u003cp\u003eLysophospholipids exhibited a distinct pattern linked to insulin regulation. We observed that several lysophosphatidylcholines and lysophosphatidylethanolamines (LPCs and LPEs) decreased after glucose ingestion in the NGT and IGT groups, whereas the T2D group only showed subtle changes in lysophospholipid levels. Instead, the T2D group showed a reduction in certain precursors like glycerol-3-phosphate. This response differs from that reported in some healthy European-ancestry populations, where postprandial LPC levels tend to rise in response to a glucose challenge [58]. The drop in lysophospholipids among the Black South African women with NGT/IGT may indicate a healthy metabolic flexibility or tissue uptake of these lipids in response to insulin. Prior work in this cohort noted that lower fasting LPC levels were associated with future T2D development [19], indicating that inadequate lysophospholipid availability could drive compensatory hypersecretion of insulin [59]. Consistent with their dual roles, we found that at 30 min post-load, higher levels of several lysophospholipids correlated with higher β-cell function, suggesting that lysophospholipids may acutely support insulin secretion, possibly by serving as signalling molecules or fuel substrates for β-cells. However, by 120 min, only a couple of lipid mediators remained significantly associated with insulin dynamics: for example, LPE (16:0) was positively associated with β-cell function, and platelet-activating factor (PAF 18:1, a specialised phospholipid) was positively associated with insulin sensitivity. These time-dependent associations suggest that lysophospholipid metabolism is linked to both insulin secretion and insulin sensitivity. Higher levels of certain lysophospholipids may support the early postprandial β-cell response, whereas later, other lipids appear more related to insulin action in peripheral tissues. The generally blunted lysophospholipid response in T2D (compared with NGT/IGT) indicates disturbed lipid handling, which may both reflect and promote β-cell dysfunction and insulin resistance in this group, highlighting these lipids as potential early biomarkers of T2D in this population.\u003c/p\u003e \u003cp\u003eCollectively, the post-OGTT metabolite responses, notably bile acids, BCAA-derivatives, carbohydrates and lysophospholipids, provide an integrated description of variations in β-cell function and insulin sensitivity. Unlike fasting metabolomics, this dynamic approach captures perturbations specifically related to glucose handling, where defects in insulin action or release manifest as altered metabolite excursions that are not evident at baseline. For example, the attenuated suppression of fatty acids and BCAAs in T2D reflects impaired insulin action despite similar fasting levels, underscoring the added value of post-challenge profiling.\u003c/p\u003e \u003cp\u003eThus, quantifying the magnitude and direction of changes in metabolites such as BCAAs or bile acids after glucose loading could help distinguish predominant defects in insulin sensitivity versus β-cell function, with potential implications for personalised interventions. Moreover, associations between specific metabolite responses (e.g. bile acids) and more favourable insulin sensitivity and secretion highlight candidate pathways for therapeutic modulation. Several limitations should be noted. While the study provides novel insights into dynamic metabolic response to glucose ingestion across glycaemic groups, the exploratory nature of non-targeted metabolomics and the relatively small sample size, highlight the need for future studies to replicate and extend these analyses in larger, independent cohorts and to evaluate their potential utility as biomarkers of T2D. Similar to other large sample size studies such as those by Liu et al., 2023 [27], future studies should consider multiple sampling across the OGTT to further characterise the temporal dynamics of metabolite responses in African populations. Additionally, while the study focuses on Black South African women, the sample may not represent the full ethnic and genetic diversity of Black African women, potentially limiting generalisability. Finally, the cross-sectional design restricts the ability to track metabolic changes over time, particularly in relation to T2D progression.\u003c/p\u003e \u003cp\u003eThe population context is indeed important. Black South African women often present with hyperinsulinaemia [59], and the distinct post-glucose lipid and amino acid profiles observed here may reflect population-specific adaptive or maladaptive mechanisms. These data emphasise that T2D pathophysiology cannot be fully captured by uniform models and that population-specific metabolic profiling is essential. By linking dynamic metabolite signatures to insulin sensitivity and secretion in this high-risk group, our study identifies potential early markers of T2D risk and motivates longitudinal and interventional studies to test their predictive and mechanistic relevance.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis study highlights the value of dynamic metabolomic profiling during an OGTT in uncovering early metabolic disturbances linked to insulin resistance and β-cell dysfunction in Black SA women. Shifts in bile acids, lysophospholipids, BCCAs, fatty acids and acylcarnitines following an OGTT revealed distinct metabolic signatures associated with glycaemic status and insulin dynamics, observations that are not apparent in the fasting state. These findings suggest that specific metabolite responses to glucose may serve as early biomarkers of T2D progression and development and offer potential targets for intervention. Future studies should focus on longitudinal designs to monitor the changes in metabolic signatures over time and their role in T2D development and progression.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u0026beta;-cell, Beta-cell; \u0026beta;-CGS, Beta-cell glucose sensitivity; BCAAs, Branched-chain amino acids; BMI, Body mass index; DXA, Dual-energy X-ray absorptiometry; FM, Fat mass; FXR, Farnesoid X receptor; GC-TOF/MS, Gas chromatography time-of-flight mass spectrometry; HC, Hip circumference; IFG, Impaired fasting glucose; IGT, Impaired glucose tolerance; ISR, Insulin secretion rates; JNK, c-Jun N-terminal kinase; LC-TOF/MS, Liquid chromatography time-of-flight mass spectrometry; LPC, Lysophosphatidylcholine; LPE, Lysophosphatidylethanolamine; MASC, Middle-Aged Soweto Cohort; \u0026nbsp;NGT, Normal glucose tolerance; \u0026nbsp;OGTT, Oral glucose tolerance test; OPLS, Orthogonal partial least squares; OPLS-EP, OPLS-effect Projection; PAF, \u0026nbsp;Platelet-activating factor; PCA, Principal component analysis; PE, Phosphatidylethanolamine; SAT, Subcutaneous adipose tissue; TGR, Takeda G protein-coupled receptor 5; \u0026nbsp;VAT, Visceral adipose tissue; \u0026nbsp;VIP, Variable importance in projection; WC, Waist circumference; WHR, Waist-to-hip ratio\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our sincere gratitude to all the women in the study for their invaluable participation and contribution to this study. We also extend our appreciation to the research team for their unwavering dedication, and the Swedish Metabolomics Centre (http://www.swedishmetabolomicscentre.se/) is acknowledged for access to instrumentation and technical support. We are also grateful to the South African Medical Research Council (SAMRC)/University of the Witwatersrand (WITS) Developmental Pathways for Health Research Unit (DPHRU) at the Chris Hani Baragwanath Hospital in Soweto, Johannesburg, SA, for providing the resources and support necessary for the successful completion of this work.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contribution statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eE.C., J.H.G., L.K.M., and T.O. contributed to the conception and design of the study. Y.Z. and E.C. carried out metabolomics data analysis and the development of the corresponding methodology. L.M. performed the clinical data analysis and drafted the initial version of the manuscript. A.M. was responsible for data collection, sample processing, and clinical data analysis, and critically reviewed and edited the manuscript. K.M.U. conducted the Mari Modelling. J.H.G., L.K.M., M.F, E.C, Y.Z, K.M.U and T.O. provided critical review and revisions of the manuscript. E.C. is the guarantor of this work, with full access to all the data included in the study and assumes responsibility for the integrity and accuracy of the data and analyses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatement of Ethics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was approved by the Human Research Ethics Committee (Medical) of the University of the Witwatersrand (M150530). The procedures and risks associated with the study were explained to the participants and they all provided signed informed consent prior to participation in the study. All testing procedures were performed at the South African Medical Research Council (SAMRC)/University of the Witwatersrand Developmental Pathways for Health Research Unit (DPHRU) at the Chris Hani Baragwanath Hospital in Soweto Johannesburg.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was supported by the Swedish Research Council (Swedish Development Grant, DNR: 2014-2522) and the National Research Foundation (NRF) of SA, which provided a scholarship to Asanda Mtintsilana (grant number 111308).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOriginal data generated and analysed during this study are available in a public repository. All processed metabolomics data, sample metadata, and metabolomics annotations have been deposited in Zenodo (doi:10.5281/zenodo.16977693). During peer review, access is provided to editors and reviewers via a private link; upon publication the repository will be released publicly under CC BY 4.0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatements and Declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no competing interests to declare that are relevant to the content of this article.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003ePheiffer C, Wyk VP Van, Turawa E, Levitt N, Kengne AP, Bradshaw D (2021) Prevalence of type 2 diabetes in South Africa: A systematic review and meta-analysis. Int J Environ Res Public Health 18(11):5868. https://doi.org/10.3390/IJERPH18115868/S1\u003c/li\u003e\n\u003cli\u003eInternational Diabetes Federation. IDF Atlas 11th Edition (2025) IDF Diabetes Atlas, 11th Edition. https://diabetesatlas.org/media/uploads/sites/3/2025/04/IDF_Atlas_11th_Edition_2025-1.pdf. Accessed 27 Jan 2026\u003c/li\u003e\n\u003cli\u003eStatistics South Africa (2020) STATISTICAL RELEASE Mortality and causes of death in South Africa: Findings from death notification 2020. www.statssa.gov.za,[email protected],Tel+27123108911\u003c/li\u003e\n\u003cli\u003eHasson BR, Apovian C, Istfan N (2015) Racial/Ethnic Differences in Insulin Resistance and Beta Cell Function: Relationship to Racial Disparities in Type 2 Diabetes among African Americans versus Caucasians. Curr Obes Rep 4(2):241\u0026ndash;249. https://doi.org/10.1007/S13679-015-0150-2/METRICS\u003c/li\u003e\n\u003cli\u003eGoedecke JH, Olsson T (2020) Pathogenesis of type 2 diabetes risk in black Africans: a South African perspective. J Intern Med 288(3):284\u0026ndash;294. https://doi.org/10.1111/JOIM.13083\u003c/li\u003e\n\u003cli\u003eGoedecke JH, Mendham AE (2022) Pathophysiology of type 2 diabetes in sub-Saharan Africans. Diabetologia 65(12):1967. https://doi.org/10.1007/S00125-022-05795-2\u003c/li\u003e\n\u003cli\u003eLadwa M, Bello O, Hakim O, et al (2020) Insulin clearance as the major player in the hyperinsulinaemia of black African men without diabetes. Diabetes Obes Metab 22(10):1808\u0026ndash;1817. https://doi.org/10.1111/DOM.14101\u003c/li\u003e\n\u003cli\u003eEsser N, Utzschneider KM, Kahn SE (2020) Early beta cell dysfunction vs insulin hypersecretion as the primary event in the pathogenesis of dysglycaemia. Diabetologia 63(10):2007\u0026ndash;2021. https://doi.org/10.1007/s00125-020-05245-x\u003c/li\u003e\n\u003cli\u003eTric\u0026ograve; D, Natali A, Arslanian S, Mari A, Ferrannini E (2018) Identification, pathophysiology, and clinical implications of primary insulin hypersecretion in nondiabetic adults and adolescents. JCI Insight 3(24). https://doi.org/10.1172/jci.insight.124912\u003c/li\u003e\n\u003cli\u003eBergman RN, Piccinini F, Kabir M, Kolka CM, Ader M (2019) Hypothesis: Role of Reduced Hepatic Insulin Clearance in the Pathogenesis of Type 2 Diabetes. Diabetes 68(9):1709\u0026ndash;1716. https://doi.org/10.2337/DB19-0098\u003c/li\u003e\n\u003cli\u003eGoedecke JH, Hellig J, Conradie-Smit M, May W (2025) Epidemiology of adult obesity. S Afr Med J 115(9b):16\u0026ndash;25. https://doi.org/10.7196/SAMJ.2025.v115i9b.3616\u003c/li\u003e\n\u003cli\u003eShi L, Brunius C, Lehtonen M, et al (2018) Plasma metabolites associated with type 2 diabetes in a Swedish population: a case\u0026ndash;control study nested in a prospective cohort. Diabetologia 61(4):849\u0026ndash;861. https://doi.org/10.1007/S00125-017-4521-Y/FIG.S/3\u003c/li\u003e\n\u003cli\u003eAhola-Olli A V., Mustelin L, Kalimeri M, et al (2019) Circulating metabolites and the risk of type 2 diabetes: a prospective study of 11,896 young adults from four Finnish cohorts. Diabetologia 62(12):2298\u0026ndash;2309. https://doi.org/10.1007/S00125-019-05001-W/TABLES/2\u003c/li\u003e\n\u003cli\u003eUrpi-Sarda M, Almanza-Aguilera E, Llorach R, et al (2019) Non-targeted metabolomic biomarkers and metabotypes of type 2 diabetes: A cross-sectional study of PREDIMED trial participants. Diabetes Metab 45(2):167\u0026ndash;174. https://doi.org/10.1016/J.DIABET.2018.02.006\u003c/li\u003e\n\u003cli\u003eLu Y, Wang Y, Ong CN, et al (2016) Metabolic signatures and risk of type 2 diabetes in a Chinese population: an untargeted metabolomics study using both LC-MS and GC-MS. Diabetologia 59(11):2349\u0026ndash;2359. https://doi.org/10.1007/S00125-016-4069-2/FIG.S/4\u003c/li\u003e\n\u003cli\u003eMorze J, Wittenbecher C, Schwingshackl L, et al (2022) Metabolomics and Type 2 Diabetes Risk: An Updated Systematic Review and Meta-analysis of Prospective Cohort Studies. Diabetes Care 45(4):1013\u0026ndash;1024. https://doi.org/10.2337/dc21-1705\u003c/li\u003e\n\u003cli\u003eChoucair I, Mallela DP, Hilser JR, et al (2024) Comprehensive Clinical and Genetic Analyses of Circulating Bile Acids and Their Associations With Diabetes and Its Indices. Diabetes 73(8):1215\u0026ndash;1228. https://doi.org/10.2337/DB23-0676\u003c/li\u003e\n\u003cli\u003eJiang X, Zhu F, Gra\u0026ccedil;a G, et al (2025) Serum Metabolomic Profiling of Incident Type 2 Diabetes Mellitus in the Multi-ethnic Study of Atherosclerosis and Rotterdam Study. J Clin Endocrinol Metab 110(8):e2700\u0026ndash;e2710. https://doi.org/10.1210/CLINEM/DGAE812\u003c/li\u003e\n\u003cli\u003eZeng Y, Mtintsilana A, Goedecke JH, Micklesfield LK, Olsson T, Chorell E (2019) 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 95:57\u0026ndash;64. https://doi.org/10.1016/j.metabol.2019.04.001\u003c/li\u003e\n\u003cli\u003eDoumatey AP, Shriner D, Zhou J, et al (2024) Untargeted metabolomic profiling reveals molecular signatures associated with type 2 diabetes in Nigerians. Genome Med 16(1):1\u0026ndash;18. https://doi.org/10.1186/S13073-024-01308-5/FIG.S/7\u003c/li\u003e\n\u003cli\u003eYang J, Chi Y, Burkhardt BR, Guan Y, Wolf BA (2010) Leucine metabolism in regulation of insulin secretion from pancreatic beta cells. Nutr Rev 68(5):270\u0026ndash;279. https://doi.org/10.1111/J.1753-4887.2010.00282.X\u003c/li\u003e\n\u003cli\u003eYea K, Kim J, Yoon JH, et al (2009) Lysophosphatidylcholine activates adipocyte glucose uptake and lowers blood glucose levels in murine models of diabetes. Journal of Biological Chemistry 284(49):33833\u0026ndash;33840. https://doi.org/10.1074/jbc.M109.024869\u003c/li\u003e\n\u003cli\u003eHan MS, Lim YM, Quan W, et al (2011) Lysophosphatidylcholine as an effector of fatty acid-induced insulin resistance. J Lipid Res 52(6):1234\u0026ndash;1246. https://doi.org/10.1194/JLR.M014787\u003c/li\u003e\n\u003cli\u003eDrzazga A, Kristinsson H, Sałaga M, et al (2018) Lysophosphatidylcholine and its phosphorothioate analogues potentiate insulin secretion via GPR40 (FFAR1), GPR55 and GPR119 receptors in a different manner. Mol Cell Endocrinol 472:117\u0026ndash;125. https://doi.org/10.1016/J.MCE.2017.12.002\u003c/li\u003e\n\u003cli\u003eWang X-H, Xu F, Cheng M, et al Fasting serum total bile acid levels are associated with insulin sensitivity, islet \u0026beta;-cell function and glucagon levels in response to glucose challenge in patients with. jstage.jst.go.jpX Wang, F Xu, M Cheng, X Wang, D Zhang, L Zhao, H Cai, H Huang, T Chen, X ZhangEndocrine Journal, 2020\u0026bull;jstage.jst.go.jp 2020(11):1107\u0026ndash;1117\u003c/li\u003e\n\u003cli\u003eGuasch-Ferr\u0026eacute; M, Hruby A, Toledo E, et al (2016) Metabolomics in Prediabetes and Diabetes: A Systematic Review and Meta-analysis. Diabetes Care 39(5):833\u0026ndash;846. https://doi.org/10.2337/DC15-2251\u003c/li\u003e\n\u003cli\u003eLiu Y, Liu X, He Q, Huang X, Ren Y, Dong Z (2024) Changes in Isoleucine, Sarcosine, and Dimethylglycine During OGTT as Risk Factors for Diabetes. J Clin Endocrinol Metab 109(7):1793\u0026ndash;1802. https://doi.org/10.1210/CLINEM/DGAE018\u003c/li\u003e\n\u003cli\u003eGoedecke JH, Nguyen KA, Kufe C, et al (2022) Waist circumference thresholds predicting incident dysglycaemia and type 2 diabetes in Black African men and women. Diabetes Obes Metab 24(5):918\u0026ndash;927. https://doi.org/10.1111/dom.14655\u003c/li\u003e\n\u003cli\u003eMtintsilana A, Micklesfield LK, Chorell E, Olsson T, Goedecke JH (2019) Fat redistribution and accumulation of visceral adipose tissue predicts type 2 diabetes risk in middle-aged black South African women: a 13-year longitudinal study. Nutr Diabetes 9(1):12. https://doi.org/10.1038/S41387-019-0079-8\u003c/li\u003e\n\u003cli\u003eMicklesfield LK, Westgate K, Smith A, et al (2022) Physical Activity Behaviors of a Middle-Age South African Cohort as Determined by Integrated Hip and Thigh Accelerometry. Med Sci Sports Exerc 54(9):1493\u0026ndash;1505. https://doi.org/10.1249/MSS.0000000000002940\u003c/li\u003e\n\u003cli\u003eWorld Health Organization (2000) Obesity:preventing and managing the global epidemic. Geneva, Switzerland\u003c/li\u003e\n\u003cli\u003eMicklesfield LK, Goedecke JH, Punyanitya M, Wilson KE, Kelly TL (2012) Dual-Energy X-Ray Performs as Well as Clinical Computed Tomography for the Measurement of Visceral Fat. Obesity 20(5):1109\u0026ndash;1114. https://doi.org/10.1038/OBY.2011.367\u003c/li\u003e\n\u003cli\u003eBalkau B, Charles MA (1999) Comment on the provisional report from the WHO consultation. Diabetic Medicine 16(5):442\u0026ndash;443. https://doi.org/10.1046/j.1464-5491.1999.00059.x\u003c/li\u003e\n\u003cli\u003eMari A, Ferrannini E (2008) \u0026beta;-cell function assessment from modelling of oral tests: an effective approach. Diabetes Obes Metab 10(SUPPL. 4):77\u0026ndash;87. https://doi.org/10.1111/J.1463-1326.2008.00946.X\u003c/li\u003e\n\u003cli\u003eMari A, Schmitz O, Gastaldelli A, Oestergaard T, Nyholm B, Ferrannini E (2002) Meal and oral glucose tests for assessment of beta -cell function: modeling analysis in normal subjects. Am J Physiol Endocrinol Metab 283(6). https://doi.org/10.1152/ajpendo.00093.2002\u003c/li\u003e\n\u003cli\u003eVan Cauter E, Mestrez F, Sturis J, Polonsky KS (1992) Estimation of insulin secretion rates from C-peptide levels. Comparison of individual and standard kinetic parameters for C-peptide clearance. Diabetes 41(3):368\u0026ndash;377. https://doi.org/10.2337/DIAB.41.3.368\u003c/li\u003e\n\u003cli\u003eMari A, Tura A, Grespan E, Bizzotto R (2020) Mathematical Modeling for the Physiological and Clinical Investigation of Glucose Homeostasis and Diabetes. Front Physiol 11:575789. https://doi.org/10.3389/FPHYS.2020.575789/XML/NLM\u003c/li\u003e\n\u003cli\u003eMatsuda M, DeFronzo RA (1999) Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22(9):1462\u0026ndash;1470. https://doi.org/10.2337/DIACARE.22.9.1462\u003c/li\u003e\n\u003cli\u003eJiye A, Trygg J, Gullberg J, et al (2005) Extraction and GC/MS analysis of the human blood plasma metabolome. Anal Chem 77(24):8086\u0026ndash;8094. https://doi.org/10.1021/ac051211v\u003c/li\u003e\n\u003cli\u003eJonsson P, Wuolikainen A, Thysell E, et al (2015) Constrained randomization and multivariate effect projections improve information extraction and biomarker pattern discovery in metabolomics studies involving dependent samples. Metabolomics 11(6):1667\u0026ndash;1678. https://doi.org/10.1007/S11306-015-0818-3/FIG.S/5\u003c/li\u003e\n\u003cli\u003eKim H, Fang S (2018) Crosstalk between FXR and TGR5 controls glucagon-like peptide 1 secretion to maintain glycemic homeostasis. Lab Anim Res 34(4):140\u0026ndash;146. https://doi.org/10.5625/lar.2018.34.4.140\u003c/li\u003e\n\u003cli\u003eHansen M, Sonne DP, Mikkelsen KH, Gluud LL, Vilsb\u0026oslash;ll T, Knop FK (2017) Bile acid sequestrants for glycemic control in patients with type 2 diabetes: A systematic review with meta-analysis of randomized controlled trials. J Diabetes Complications 31(5):918\u0026ndash;927. https://doi.org/10.1016/J.JDIACOMP.2017.01.011\u003c/li\u003e\n\u003cli\u003eCadena Sandoval M, Haeusler RA (2025) Bile acid metabolism in type 2 diabetes mellitus. Nat Rev Endocrinol 21(4):203\u0026ndash;213. https://doi.org/10.1038/S41574-024-01067-8;SUBJMETA\u003c/li\u003e\n\u003cli\u003eTicho AL, Malhotra P, Dudeja PK, Gill RK, Alrefai WA (2019) Bile acid receptors and gastrointestinal functions. Liver Res 3(1):31\u0026ndash;39. https://doi.org/10.1016/j.livres.2019.01.001\u003c/li\u003e\n\u003cli\u003eD\u0026uuml;fer M, H\u0026ouml;rth K, Wagner R, et al (2012) Bile acids acutely stimulate insulin secretion of mouse \u0026beta;-cells via farnesoid X receptor activation and K(ATP) channel inhibition. Diabetes 61(6):1479\u0026ndash;1489. https://doi.org/10.2337/DB11-0815\u003c/li\u003e\n\u003cli\u003eNewgard CB, An J, Bain JR, et al (2009) A Branched-Chain Amino Acid-Related Metabolic Signature that Differentiates Obese and Lean Humans and Contributes to Insulin Resistance. Cell Metab 9(4):311\u0026ndash;326. https://doi.org/10.1016/J.CMET.2009.02.002/ATTACHMENT/2556051D-632E-4829-BEF5-65236D548452/MMC1.PDF\u003c/li\u003e\n\u003cli\u003eHo JE, Larson MG, Vasan RS, et al (2013) Metabolite profiles during oral glucose challenge. Diabetes 62(8):2689\u0026ndash;2698. https://doi.org/10.2337/DB12-0754/-/DC1\u003c/li\u003e\n\u003cli\u003eSj\u0026ouml;gren RJO, Rizo-Roca D, Chibalin A V., et al (2021) Branched-chain amino acid metabolism is regulated by ERR\u0026alpha; in primary human myotubes and is further impaired by glucose loading in type 2 diabetes. Diabetologia 64(9):2077\u0026ndash;2091. https://doi.org/10.1007/S00125-021-05481-9\u003c/li\u003e\n\u003cli\u003eFanni G, Eriksson JW, Pereira MJ (2023) Several Metabolite Families Display Inflexibility during Glucose Challenge in Patients with Type 2 Diabetes: An Untargeted Metabolomics Study. Metabolites 13(1):131. https://doi.org/10.3390/METABO13010131/S1\u003c/li\u003e\n\u003cli\u003eGannon NP, Schnuck JK, Vaughan RA (2018) BCAA Metabolism and Insulin Sensitivity \u0026ndash; Dysregulated by Metabolic Status? Mol Nutr Food Res 62(6):1700756. https://doi.org/10.1002/MNFR.201700756\u003c/li\u003e\n\u003cli\u003eAdams SH (2011) Emerging Perspectives on Essential Amino Acid Metabolism in Obesity and the Insulin-Resistant State. Advances in Nutrition 2(6):445. https://doi.org/10.3945/AN.111.000737\u003c/li\u003e\n\u003cli\u003eKrebs M, Krssak M, Bernroider E, et al (2002) Mechanism of amino acid-induced skeletal muscle insulin resistance in humans. Diabetes 51(3):599\u0026ndash;605. https://doi.org/10.2337/DIABETES.51.3.599\u003c/li\u003e\n\u003cli\u003eFerrannini E, Natali A, Camastra S, et al (2013) Early metabolic markers of the development of dysglycemia and type 2 diabetes and their physiological significance. Diabetes 62(5):1730\u0026ndash;1737. https://doi.org/10.2337/DB12-0707/-/DC1\u003c/li\u003e\n\u003cli\u003eLynch CJ, Adams SH (2014) Branched-chain amino acids in metabolic signalling and insulin resistance. Nat Rev Endocrinol 10(12):723. https://doi.org/10.1038/NRENDO.2014.171\u003c/li\u003e\n\u003cli\u003eKim MJ, Jung HS, Hwang-Bo Y, et al (2013) Evaluation of 1,5-anhydroglucitol as a marker for glycemic variability in patients with type 2 diabetes mellitus. Acta Diabetol 50(4):505\u0026ndash;510. https://doi.org/10.1007/S00592-011-0302-0\u003c/li\u003e\n\u003cli\u003eMa H, Li X, Zhou T, et al (2020) Glucosamine Use, Inflammation, and Genetic Susceptibility, and Incidence of Type 2 Diabetes: A Prospective Study in UK Biobank. Diabetes Care 43(4):719\u0026ndash;725. https://doi.org/10.2337/DC19-1836\u003c/li\u003e\n\u003cli\u003eZhou S, Zhou P, Yang T, Si J, An W, Jiang Y (2025) Glucosamine supplementation contributes to reducing the risk of type 2 diabetes: Evidence from Mendelian randomization combined with a meta-analysis. J Int Med Res 53(4):03000605251334460. https://doi.org/10.1177/03000605251334460\u003c/li\u003e\n\u003cli\u003eZhao X, Peter A, Fritsche J, et al (2009) Changes of the plasma metabolome during an oral glucose tolerance test: Is there more than glucose to look at? Am J Physiol Endocrinol Metab 296(2):384\u0026ndash;393. https://doi.org/10.1152/AJPENDO.90748.2008/ASSET/IMAGES/LARGE/ZH10020955660008.JPEG\u003c/li\u003e\n\u003cli\u003eGoedecke JH, Dave JA, Faulenbach M V., et al (2009) Insulin response in relation to insulin sensitivity: an appropriate beta-cell response in black South African women. Diabetes Care 32(5):860\u0026ndash;865. https://doi.org/10.2337/dc08-2048 \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"metabologia","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Metabologia](https://link.springer.com/journal/44357)","snPcode":"44357","submissionUrl":"https://submission.springernature.com/new-submission/44357/3?","title":"Metabologia","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Open","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Type 2 diabetes, insulin sensitivity, beta-cell function, metabolomics, Black women, South Africa","lastPublishedDoi":"10.21203/rs.3.rs-9160380/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9160380/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003ePostprandial metabolic responses are strong predictors of type 2 diabetes (T2D) and its underlying pathophysiological traits, insulin sensitivity and β-cell function, but remain poorly characterised in African populations. We investigated glucose-stimulated metabolite profiles across the glycaemic spectrum and their associations with insulin sensitivity and β-cell function in Black South African women.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis cross-sectional study included 65 women (median age: 56 years) from the Middle-Aged Soweto Cohort, categorised as normal glucose tolerance (NGT, n\u0026thinsp;=\u0026thinsp;29), impaired glucose tolerance (IGT, n\u0026thinsp;=\u0026thinsp;24) and T2D (n\u0026thinsp;=\u0026thinsp;12). Following an oral glucose tolerance test, insulin sensitivity (Matsuda index) and β-cell function were estimated from the Mari model. Changes in metabolic profiles were characterised from fasting to 30 (Δ30) and 120 (Δ120) minutes post-glucose ingestion using a multi-platform metabolomics approach.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eInsulin sensitivity and β-cell function declined progressively from NGT to IGT to T2D. From Δ30 and Δ120, carbohydrates and bile acids increased, whereas amino acids (including BCAAs), fatty acids and lysophospholipids decreased across all groups; and associated with insulin sensitivity and β-cell function. At Δ120, bile acids, cholic acid and deoxycholic acid, remained elevated in the T2D group only. Lysophospholipids decreased across groups. Carbohydrates, bile acids, and lysophospholipids correlated positively with insulin sensitivity and β-cell function, while amino acids, BCAAs, and fatty acids correlated negatively.\u003c/p\u003e\u003ch2\u003eConclusion:\u003c/h2\u003e \u003cp\u003eDistinct post-glucose metabolite responses across the glycaemic spectrum reflect differences in insulin sensitivity and β-cell function in Black South African women, highlighting the value of dynamic metabolic profiling for understanding T2D progression in African populations.\u003c/p\u003e","manuscriptTitle":"Post-glucose Metabolite Signatures Reflect Insulin Sensitivity and Beta-cell Function in Black South African Women","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-07 15:26:30","doi":"10.21203/rs.3.rs-9160380/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-01T16:07:53+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-21T05:41:04+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-19T16:19:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"246150428601379200666500126729307703490","date":"2026-04-14T07:02:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"30435619768659944983604424374490670361","date":"2026-04-14T04:47:29+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-01T12:27:30+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-31T13:10:25+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-31T11:43:10+00:00","index":"","fulltext":""},{"type":"submitted","content":"Metabologia","date":"2026-03-25T14:23:22+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"metabologia","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Metabologia](https://link.springer.com/journal/44357)","snPcode":"44357","submissionUrl":"https://submission.springernature.com/new-submission/44357/3?","title":"Metabologia","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Open","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"be29e130-74ad-4918-8e1e-86ced20adf4d","owner":[],"postedDate":"April 7th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-01T16:07:53+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-01T16:23:33+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-07 15:26:30","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9160380","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9160380","identity":"rs-9160380","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}